improve verify_sink_dims

2025-03-18 19:38:22 +08:00
376 changed files with 243032 additions and 1337149 deletions
@@ -121,8 +121,9 @@ runs:
        echo -e 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
        sudo apt update || true
        sudo apt install --no-install-recommends --allow-unauthenticated -y hsa-rocr comgr hsa-rocr-dev liburing-dev libc6-dev
-        cargo build --release --manifest-path ./extra/remu/Cargo.toml
-        sudo ln -sf ${{ github.workspace }}/extra/remu/target/release/libremu.so /usr/local/lib/libremu.so
+        curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
+        jq -r '.assets[] | select(.name == "libremu.so").browser_download_url' | \
+        sudo xargs curl -L -o /usr/local/lib/libremu.so
        sudo tee --append /etc/ld.so.conf.d/rocm.conf <<'EOF'
          /opt/rocm/lib
          /opt/rocm/lib64
@@ -136,7 +137,9 @@ runs:
        curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/nimlgen/amdcomgr_dylib/releases/latest | \
          jq -r '.assets[] | select(.name == "libamd_comgr.dylib").browser_download_url' | \
          sudo xargs curl -L -o /usr/local/lib/libamd_comgr.dylib
-        cargo build --release --manifest-path ./extra/remu/Cargo.toml
+        curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
+          jq -r '.assets[] | select(.name == "libremu.dylib").browser_download_url' | \
+          sudo xargs curl -L -o /usr/local/lib/libremu.dylib

    # **** CUDA ****

@@ -172,7 +175,7 @@ runs:
        git checkout b16039dc940dc6bc4ea0a98380495769ff35ed99
        mkdir build
        cd build
-        cmake .. -Wno-dev -G Ninja -DOCELOT_BUILD_TOOLS=OFF -DCMAKE_BUILD_ALWAYS=0 -DBUILD_TESTS_CUDA=OFF -DCMAKE_POLICY_VERSION_MINIMUM=3.5
+        cmake .. -Wno-dev -G Ninja -DOCELOT_BUILD_TOOLS=OFF -DCMAKE_BUILD_ALWAYS=0 -DBUILD_TESTS_CUDA=OFF
        ninja
    - name: Install gpuocelot
      if: inputs.cuda == 'true'
@@ -199,8 +202,8 @@ runs:
      if: inputs.webgpu == 'true' && runner.os == 'macOS'
      shell: bash
      run: |
-        brew tap wpmed92/dawn
-        brew install dawn
+        sudo mkdir -p /usr/local/lib
+        sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.dylib -o /usr/local/lib/libwebgpu_dawn.dylib

    # **** LLVM ****

@@ -209,10 +212,8 @@ runs:
      shell: bash
      run: |
        echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
-        wget -qO- https://apt.llvm.org/llvm-snapshot.gpg.key | sudo tee /etc/apt/trusted.gpg.d/apt.llvm.org.asc
-        echo "deb http://apt.llvm.org/$(lsb_release -cs)/ llvm-toolchain-$(lsb_release -cs)-19 main" | sudo tee /etc/apt/sources.list.d/llvm.list
        sudo apt update -y || true
-        sudo apt install -y --no-install-recommends libllvm19 clang-19 lld-19
+        sudo apt install -y --no-install-recommends llvm-dev

    - name: Install LLVM (macOS)
      if: inputs.llvm == 'true' && runner.os == 'macOS'
@@ -61,17 +61,17 @@ jobs:
    - name: Test speed vs torch
      run: BIG=2 MPS=1 python3.11 test/test_speed_v_torch.py | tee torch_speed.txt
    - name: Test tensor cores
-      run: METAL=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+      run: METAL=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Test AMX tensor cores
      run: |
-        DEBUG=2 CPU=1 AMX=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
-        DEBUG=2 LLVM=1 AMX=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+        DEBUG=2 CPU=1 AMX=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
+        DEBUG=2 LLVM=1 AMX=1 python3.11 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Run Tensor Core GEMM (float)
-      run: DEBUG=2 SHOULD_USE_TC=1 python3.11 extra/gemm/simple_matmul.py | tee matmul.txt
+      run: DEBUG=2 python3.11 extra/gemm/simple_matmul.py | tee matmul.txt
    - name: Run Tensor Core GEMM (half)
-      run: DEBUG=2 SHOULD_USE_TC=1 HALF=1 python3.11 extra/gemm/simple_matmul.py | tee matmul_half.txt
+      run: DEBUG=2 HALF=1 python3.11 extra/gemm/simple_matmul.py | tee matmul_half.txt
    - name: Run Tensor Core GEMM (bfloat16)
-      run: DEBUG=2 SHOULD_USE_TC=1 BFLOAT16=1 python3.11 extra/gemm/simple_matmul.py | tee matmul_bfloat16.txt
+      run: DEBUG=2 BFLOAT16=1 python3.11 extra/gemm/simple_matmul.py | tee matmul_bfloat16.txt
    - name: Fuzz Padded Tensor Core GEMM
      run: METAL=1 M_START=6 M_STOP=10 M_STEP=1 N_START=6 N_STOP=10 N_STEP=1 K_START=6 K_STOP=24 K_STEP=1 TC_OPT=2 DEBUG=2 python3.11 ./extra/gemm/fuzz_matmul.py
    - name: Run LLaMA
@@ -174,17 +174,17 @@ jobs:
      run: NV=1 python test/external/external_benchmark_multitensor_allreduce.py
    - name: Test tensor cores
      run: |
-        NV=1 ALLOW_TF32=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
-        PTX=1 ALLOW_TF32=1 NV=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+        NV=1 ALLOW_TF32=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
+        PTX=1 ALLOW_TF32=1 NV=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Run Tensor Core GEMM (CUDA)
      run: |
-        CUDA=1 SHOULD_USE_TC=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul.txt
-        CUDA=1 SHOULD_USE_TC=1 BFLOAT16=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_bfloat16.txt
-        CUDA=1 SHOULD_USE_TC=1 ALLOW_TF32=1 DEBUG=2 ATOL=2e-2 python3 extra/gemm/simple_matmul.py | tee matmul_tf32.txt
+        CUDA=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul.txt
+        CUDA=1 BFLOAT16=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_bfloat16.txt
+        CUDA=1 ALLOW_TF32=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_tf32.txt
    - name: Run Tensor Core GEMM (PTX)
-      run: NV=1 PTX=1 SHOULD_USE_TC=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_ptx.txt
+      run: NV=1 PTX=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_ptx.txt
    - name: Run Tensor Core GEMM (NV)
-      run: NV=1 SHOULD_USE_TC=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_nv.txt
+      run: NV=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_nv.txt
    - name: Test NV=1
      run: DEBUG=2 NV=1 python -m pytest -rA test/test_tiny.py
    - name: Test CUDA=1
@@ -369,11 +369,9 @@ jobs:
    - name: Test speed vs theoretical
      run: AMD=1 IGNORE_BEAM_CACHE=1 BEAM_DEBUG=1 DEBUG=1 python -m pytest -rA test/external/speed_v_theoretical.py --durations=20
    - name: Test tensor cores
-      run: |
-        AMD=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded_amd TestLinearizer.test_tensor_cores_padded_uops
-        AMD=1 SHOULD_USE_TC=1 BFLOAT16=1 DEBUG=2 python3 extra/gemm/simple_matmul.py
+      run: AMD=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Run Tensor Core GEMM (AMD)
-      run: AMD=1 SHOULD_USE_TC=1 HALF=1 DEBUG=2 ATOL=2e-2 python3 extra/gemm/simple_matmul.py | tee matmul_amd.txt
+      run: AMD=1 HALF=1 DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul_amd.txt
    - name: Test AMD=1
      run: DEBUG=2 AMD=1 python -m pytest -rA test/test_tiny.py
    - name: Test HIP=1
@@ -88,7 +88,7 @@ jobs:
  autogen:
    name: Autogen
    runs-on: ubuntu-22.04
-    timeout-minutes: 15
+    timeout-minutes: 10
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -116,17 +116,14 @@ jobs:
    - name: Verify AMD autogen
      run: |
        cp tinygrad/runtime/autogen/hsa.py /tmp/hsa.py.bak
-        cp tinygrad/runtime/autogen/kfd.py /tmp/kfd.py.bak
        cp tinygrad/runtime/autogen/comgr.py /tmp/comgr.py.bak
        cp tinygrad/runtime/autogen/amd_gpu.py /tmp/amd_gpu.py.bak
        cp tinygrad/runtime/autogen/sqtt.py /tmp/sqtt.py.bak
        ./autogen_stubs.sh hsa
-        ./autogen_stubs.sh kfd
        ./autogen_stubs.sh comgr
        ./autogen_stubs.sh amd
        ./autogen_stubs.sh sqtt
        diff /tmp/hsa.py.bak tinygrad/runtime/autogen/hsa.py
-        diff /tmp/kfd.py.bak tinygrad/runtime/autogen/kfd.py
        diff /tmp/comgr.py.bak tinygrad/runtime/autogen/comgr.py
        diff /tmp/amd_gpu.py.bak tinygrad/runtime/autogen/amd_gpu.py
        diff /tmp/sqtt.py.bak tinygrad/runtime/autogen/sqtt.py
@@ -152,9 +149,7 @@ jobs:
  torchbackend:
    name: Torch Backend Tests
    runs-on: ubuntu-latest
-    timeout-minutes: 15
-    env:
-      IGNORE_OOB: 0
+    timeout-minutes: 10
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -171,7 +166,7 @@ jobs:
        sudo apt install -y --no-install-recommends ninja-build
    - name: Lint with ruff
      run: |
-        pip3 install --upgrade --force-reinstall ruff==0.11.0
+        pip3 install --upgrade --force-reinstall ruff
        python3 -m ruff check extra/torch_backend/backend.py
    - name: Test one op
      run: PYTHONPATH=. FORWARD_ONLY=1 TINY_BACKEND=1 python3 test/test_ops.py TestOps.test_add
@@ -181,8 +176,8 @@ jobs:
      run: PYTHONPATH=. python3 extra/torch_backend/test.py
    - name: Test one op in torch tests
      run: PYTHONPATH=. DEBUG=2 python3 extra/torch_backend/torch_tests.py TestTinyBackendPRIVATEUSE1.test_unary_log_tiny_float32
-    - name: Test Ops with TINY_BACKEND
-      run: PYTHONPATH=. LLVM=1 LLVMOPT=0 TINY_BACKEND=1 python3 -m pytest -n auto test/test_ops.py --durations=20
+    - name: Test Ops with TINY_BACKEND (expect failure)
+      run: PYTHONPATH=. LLVM=1 LLVMOPT=0 TINY_BACKEND=1 python3 -m pytest -n auto test/test_ops.py --durations=20 || true
    - name: Test in-place operations on views
      run: PYTHONPATH=. TORCH_DEBUG=1 python3 extra/torch_backend/test_inplace.py
    - name: Test multi-gpu
@@ -191,9 +186,7 @@ jobs:
  torchbackendmore:
    name: Torch Backend Tests More
    runs-on: ubuntu-latest
-    timeout-minutes: 15
-    env:
-      IGNORE_OOB: 0
+    timeout-minutes: 10
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -216,8 +209,6 @@ jobs:
    name: Tensor Core tests
    runs-on: ubuntu-latest
    timeout-minutes: 10
-    env:
-      IGNORE_OOB: 0
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -233,8 +224,7 @@ jobs:
    - name: Test emulated METAL tensor cores
      run: |
        DEBUG=2 EMULATE_METAL=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_big_gemm
-        PYTHONPATH=. DEBUG=2 EMULATE_METAL=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores
-        PYTHONPATH=. DEBUG=2 EMULATE_METAL=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+        PYTHONPATH=. DEBUG=2 EMULATE_METAL=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Test emulated AMX tensor cores
      run: PYTHONPATH=. DEBUG=2 AMX=1 EMULATE_AMX=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
    - name: Test emulated AMD tensor cores
@@ -243,20 +233,13 @@ jobs:
        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 N=64 HALF=1 ACC_HALF=0 python3 ./extra/gemm/simple_matmul.py
        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 N=16 HALF=1 ACC_HALF=1 python3 ./extra/gemm/simple_matmul.py
        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 N=64 HALF=1 ACC_HALF=1 python3 ./extra/gemm/simple_matmul.py
-        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores
-        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
-    - name: Test emulated AMD MFMA tensor cores
-      run: |
-        PYTHONPATH=. DEBUG=2 EMULATE_AMD_MFMA=1 FORWARD_ONLY=1 PYTHON=1 N=64 HALF=1 ACC_HALF=0 python3 ./extra/gemm/simple_matmul.py
-        PYTHONPATH=. DEBUG=2 EMULATE_AMD_MFMA=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores
-        PYTHONPATH=. DEBUG=2 EMULATE_AMD_MFMA=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+        PYTHONPATH=. DEBUG=2 EMULATE_AMD=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Test emulated CUDA tensor cores
      run: |
        DEBUG=2 EMULATE_CUDA=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm_fp16
        DEBUG=2 EMULATE_CUDA=1 ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
        DEBUG=2 EMULATE_CUDA_SM75=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm_fp16
-        PYTHONPATH="." DEBUG=2 EMULATE_CUDA=1 ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores
-        PYTHONPATH="." DEBUG=2 EMULATE_CUDA=1 ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores_padded TestLinearizer.test_tensor_cores_padded_uops
+        PYTHONPATH="." DEBUG=2 EMULATE_CUDA=1 ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
    - name: Test emulated INTEL OpenCL tensor cores
      run: DEBUG=2 EMULATE_INTEL=1 FORWARD_ONLY=1 PYTHON=1 HALF=1 N=64 python3 ./extra/gemm/simple_matmul.py
    - name: Full test tensor cores
@@ -269,10 +252,9 @@ jobs:
    - name: Test tensor cores (TC=3)
      run: |
        TC=3 DEBUG=3 EMULATE_METAL=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
-        TC=3 PYTHONPATH=. DEBUG=3 EMULATE_AMD=1 PYTHON=1 N=16 HALF=1 ACC_HALF=0 ATOL=3e-4 python3 ./extra/gemm/simple_matmul.py
-        TC=3 PYTHONPATH=. DEBUG=3 EMULATE_AMD_MFMA=1 PYTHON=1 N=16 HALF=1 ACC_HALF=0 ATOL=3e-4 python3 ./extra/gemm/simple_matmul.py
+        TC=3 PYTHONPATH=. DEBUG=3 EMULATE_AMD=1 PYTHON=1 N=16 HALF=1 ACC_HALF=0 python3 ./extra/gemm/simple_matmul.py
        TC=3 DEBUG=3 EMULATE_CUDA=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm_fp16
-        TC=3 PYTHONPATH=. DEBUG=3 EMULATE_INTEL=1 PYTHON=1 N=16 HALF=1 ATOL=3e-4 python3 ./extra/gemm/simple_matmul.py
+        TC=3 PYTHONPATH=. DEBUG=3 EMULATE_INTEL=1 PYTHON=1 N=16 HALF=1 python3 ./extra/gemm/simple_matmul.py
        TC=3 PYTHONPATH=. DEBUG=3 AMX=1 EMULATE_AMX=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
    - name: Test device flop counts
      run: |
@@ -286,8 +268,6 @@ jobs:
    name: Python Backend
    runs-on: ubuntu-latest
    timeout-minutes: 10
-    env:
-      IGNORE_OOB: 0
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -299,15 +279,13 @@ jobs:
    - name: Test dtype with Python emulator
      run: DEBUG=1 PYTHONPATH=. PYTHON=1 python3 -m pytest -n=auto test/test_dtype.py test/test_dtype_alu.py
    - name: Test ops with Python emulator
-      run: DEBUG=2 PYTHON=1 python3 -m pytest -n=auto test/test_ops.py -k "not (test_split or test_simple_cumsum or test_cumsum or test_einsum or test_dot or test_dot_1d or test_big_gemm or test_broadcastdot or test_multidot or test_var_axis or test_std_axis or test_broadcast_full or test_broadcast_partial or test_simple_conv3d or test_dilated_conv_transpose2d or test_simple_conv_transpose3d or test_large_input_conv2d or test_max_pool2d or test_max_pool2d_simple or test_max_pool2d_bigger_stride or test_avg_pool2d or test_cat or test_scaled_product_attention or test_scaled_product_attention_causal or test_slice_fancy_indexing_dim_inject_none or test_slice_fancy_indexing_list_indices or test_slice_fancy_indexing_no_dim_collapse or test_slice_fancy_indexing_tuple_indices or test_slice_fancy_indexing_list_with_tensors or test_slice_fancy_indexing_dim_collapse_int or test_interpolate_bilinear or test_interpolate_bilinear_corners_aligned or test_scaled_dot_product_attention or test_cummax or test_simple_cummax or test_logcumsumexp or test_sort or test_cumprod)" --durations=20
+      run: DEBUG=2 PYTHON=1 python3 -m pytest -n=auto test/test_ops.py -k "not (test_split or test_simple_cumsum or test_cumsum or test_einsum or test_dot or test_dot_1d or test_big_gemm or test_broadcastdot or test_multidot or test_var_axis or test_std_axis or test_broadcast_full or test_broadcast_partial or test_simple_conv3d or test_dilated_conv_transpose2d or test_simple_conv_transpose3d or test_large_input_conv2d or test_max_pool2d or test_max_pool2d_simple or test_max_pool2d_bigger_stride or test_avg_pool2d or test_cat or test_scaled_product_attention or test_scaled_product_attention_causal or test_slice_fancy_indexing_dim_inject_none or test_slice_fancy_indexing_list_indices or test_slice_fancy_indexing_no_dim_collapse or test_slice_fancy_indexing_tuple_indices or test_slice_fancy_indexing_list_with_tensors or test_slice_fancy_indexing_dim_collapse_int or test_interpolate_bilinear or test_interpolate_bilinear_corners_aligned or test_scaled_dot_product_attention or test_cummax)" --durations=20
    - name: Test uops with Python emulator
      run: PYTHON=1 python3 -m pytest test/test_uops.py --durations=20
    - name: Test symbolic with Python emulator
      run: PYTHONPATH=. PYTHON=1 python3 test/test_symbolic_ops.py
    - name: test_linearizer_failures with Python emulator
      run: PYTHONPATH=. PYTHON=1 python3 -m pytest -rA test/test_linearizer_failures.py::TestLinearizerFailures::test_failure_1
-    - name: test_renderer_failures with Python emulator
-      run: PYTHONPATH=. PYTHON=1 python3 -m pytest -rA test/test_renderer_failures.py::TestRendererFailures

  linter:
    name: Linters
@@ -328,7 +306,7 @@ jobs:
      run: python -m pylint --disable=all -e W0311 -e C0303 --jobs=0 --indent-string='  ' --recursive=y .
    - name: Lint with ruff
      run: |
-        pip3 install --upgrade --force-reinstall ruff==0.11.0
+        pip3 install --upgrade --force-reinstall ruff
        python3 -m ruff check .
    - name: Lint tinygrad with pylint
      run: python -m pylint tinygrad/
@@ -352,12 +330,8 @@ jobs:
      run: awk '/```python/{flag=1;next}/```/{flag=0}flag' README.md > README.py &&  PYTHONPATH=. python README.py
    - name: Run unit tests
      run: PYTHONPATH="." python -m pytest -n=auto test/unit/
-    - name: Run targetted tests on NULL backend
-      run: PYTHONPATH="." NULL=1 python3 test/test_multitensor.py TestMultiTensor.test_data_parallel_resnet_train_step
-    - name: Run GC tests
-      run: PYTHONPATH="." python test/external/external_uop_gc.py
-    - name: Repo line count < 12500 lines
-      run: MAX_LINE_COUNT=12500 python sz.py
+    - name: Repo line count < 12000 lines
+      run: MAX_LINE_COUNT=12000 python sz.py

  fuzzing:
    name: Fuzzing
@@ -373,8 +347,6 @@ jobs:
        deps: testing_minimal
    - name: Fuzz Test symbolic
      run: python test/external/fuzz_symbolic.py
-    - name: Fuzz Test fast idiv
-      run: python test/external/fuzz_fast_idiv.py
    - name: Fuzz Test shapetracker
      run: |
        PYTHONPATH="." python test/external/fuzz_shapetracker.py
@@ -384,8 +356,6 @@ jobs:
    name: 'GPU IMAGE Tests'
    runs-on: ubuntu-22.04
    timeout-minutes: 10
-    env:
-      IGNORE_OOB: 0
    steps:
      - name: Checkout Code
        uses: actions/checkout@v4
@@ -410,8 +380,6 @@ jobs:
    name: 'openpilot Compile Tests'
    runs-on: ubuntu-22.04
    timeout-minutes: 10
-    env:
-      IGNORE_OOB: 0
    steps:
      - name: Checkout Code
        uses: actions/checkout@v4
@@ -435,8 +403,6 @@ jobs:
    name: 'ONNX+Optimization Tests'
    runs-on: ubuntu-22.04
    timeout-minutes: 20
-    env:
-      IGNORE_OOB: 0

    steps:
      - name: Checkout Code
@@ -461,8 +427,8 @@ jobs:
        run: CPU=1 PYTHONPATH=. python3 test/test_quantize_onnx.py
      - name: Run CLOUD=1 Test
        run: |
-          CLOUDDEV=CPU CLOUD=1 python3 -m pytest test/test_tiny.py test/test_jit.py
-          CLOUDDEV=GPU CLOUD=1 python3 -m pytest test/test_tiny.py test/test_jit.py
+          CLOUDDEV=CPU CLOUD=1 python3 test/test_tiny.py
+          CLOUDDEV=GPU CLOUD=1 python3 test/test_tiny.py
          CLOUDDEV=GPU IMAGE=2 CLOUD=1 python3 test/test_tiny.py
      - name: Test Optimization Helpers
        run: PYTHONPATH="." DEBUG=1 python3 extra/optimization/test_helpers.py
@@ -481,8 +447,6 @@ jobs:
    name: Models (llvm+cpu+gpu)
    runs-on: ubuntu-22.04
    timeout-minutes: 10
-    env:
-      IGNORE_OOB: 0
    steps:
      - name: Checkout Code
        uses: actions/checkout@v4
@@ -506,8 +470,6 @@ jobs:
    name: Linux (DSP)
    runs-on: ubuntu-24.04
    timeout-minutes: 15
-    env:
-      IGNORE_OOB: 0
    steps:
    - name: Checkout Code
      uses: actions/checkout@v4
@@ -516,7 +478,7 @@ jobs:
      with:
        key: dsp-minimal
        deps: testing_minimal
-        pydeps: "onnx==1.17.0 onnxruntime pillow"
+        pydeps: "onnx==1.16.0 onnxruntime pillow"
        llvm: "true"
    - name: Set up Docker Buildx
      uses: docker/setup-buildx-action@v3
@@ -531,8 +493,6 @@ jobs:
        cache-to: type=gha,mode=min
    - name: Run test_tiny on DSP
      run: DEBUG=2 DSP=1 python test/test_tiny.py
-    - name: Test transcendentals
-      run: CC=clang-19 PYTHONPATH="." DEBUG=2 DSP=1 python test/test_transcendental.py TestTranscendentalVectorized
    - name: Test quantize onnx
      run: PYTHONPATH="." DEBUG=2 DSP=1 python3 test/test_quantize_onnx.py
    - name: Test LLVM=1 DEVECTORIZE=0
@@ -568,85 +528,15 @@ jobs:
    - name: Run process replay tests
      uses: ./.github/actions/process-replay

-  testamd:
-    strategy:
-      fail-fast: false
-      matrix:
-        backend: [amd, amdllvm]
-
-    name: Linux (${{ matrix.backend }})
-    runs-on: ubuntu-22.04
-    timeout-minutes: 20
-    env:
-      AMD: 1
-      MOCKGPU: 1
-      FORWARD_ONLY: 1
-      AMD_LLVM: ${{ matrix.backend == 'amdllvm' && '1' || matrix.backend != 'amdllvm' && '0' }}
-    steps:
-      - name: Checkout Code
-        uses: actions/checkout@v4
-      - name: Setup Environment
-        uses: ./.github/actions/setup-tinygrad
-        with:
-          key: ${{ matrix.backend }}-minimal
-          deps: testing_minimal
-          amd: 'true'
-          llvm: ${{ matrix.backend == 'amdllvm' && 'true' }}
-      - name: Check Device.DEFAULT and print some source
-        run: |
-          PYTHONPATH=${{ github.workspace }} python3 -c "from tinygrad import Device; assert Device.DEFAULT in ['AMD'], Device.DEFAULT"
-          DEBUG=5 PYTHONPATH=${{ github.workspace }} FORWARD_ONLY=1 python3 test/test_ops.py TestOps.test_add
-      - name: Run LLVM test
-        if: matrix.backend=='amdllvm'
-        run: PYTHONPATH="." python test/test_amd_llvm.py
-      - name: Run pytest (amd)
-        run: python -m pytest -n=auto test/test_ops.py test/test_dtype.py test/test_dtype_alu.py test/test_linearizer.py test/test_randomness.py test/imported/test_indexing.py test/test_hcq.py test/external/external_test_am.py --durations=20
-      - name: Run TRANSCENDENTAL math
-        run: TRANSCENDENTAL=2 python -m pytest -n=auto test/test_ops.py::TestOps::test_sin test/test_ops.py::TestOps::test_cos test/test_ops.py::TestOps::test_tan test/test_ops.py::TestOps::test_exp test/test_ops.py::TestOps::test_log --durations=20
-      - name: Run process replay tests
-        uses: ./.github/actions/process-replay
-
-  testnvidia:
-    strategy:
-      fail-fast: false
-      matrix:
-        backend: [ptx, nv]
-
-    name: Linux (${{ matrix.backend }})
-    runs-on: ubuntu-22.04
-    timeout-minutes: 20
-
-    steps:
-      - name: Checkout Code
-        uses: actions/checkout@v4
-      - name: Setup Environment
-        uses: ./.github/actions/setup-tinygrad
-        with:
-          key: ${{ matrix.backend }}-minimal
-          deps: testing_minimal
-          cuda: 'true'
-      - name: Set env
-        run: printf "${{ matrix.backend == 'PTX' && 'FORWARD_ONLY=1\nJIT=1\nOPT=2\nCUDA=1\nPTX=1\nMOCKGPU=1' || matrix.backend == 'nv' && 'NV=1\nMOCKGPU=1\nFORWARD_ONLY=1' }}" >> $GITHUB_ENV
-      - name: Check Device.DEFAULT and print some source
-        run: |
-          PYTHONPATH=${{ github.workspace }} python3 -c "from tinygrad import Device; assert Device.DEFAULT in ['CUDA','NV'], Device.DEFAULT"
-          DEBUG=5 PYTHONPATH=${{ github.workspace }} FORWARD_ONLY=1 python3 test/test_ops.py TestOps.test_add
-      - name: Run pytest (cuda)
-        run: python -m pytest -n=auto test/ --ignore=test/models --ignore=test/unit --ignore test/test_gc.py --durations=20
-      - name: Run process replay tests
-        uses: ./.github/actions/process-replay
-
  tests:
    strategy:
      fail-fast: false
      matrix:
-        backend: [llvm, cpu, gpu]
+        backend: [llvm, cpu, gpu, ptx, amd, nv] #, triton]

    name: Linux (${{ matrix.backend }})
    runs-on: ubuntu-22.04
    timeout-minutes: 20
-    env:
-      IGNORE_OOB: 0

    steps:
      - name: Checkout Code
@@ -655,17 +545,25 @@ jobs:
        uses: ./.github/actions/setup-tinygrad
        with:
          key: ${{ matrix.backend }}-minimal
-          deps: testing_minimal
+          deps: testing_minimal${{matrix.backend=='ptx'&&',cuda'||matrix.backend=='triton'&&',triton'||''}}
          opencl: ${{ matrix.backend == 'gpu' && 'true' }}
-          llvm: ${{ matrix.backend == 'llvm' && 'true' }}
+          amd: ${{ matrix.backend == 'amd' && 'true' }}
+          cuda: ${{ (matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv') && 'true' }}
      - name: Set env
-        run: printf "${{ matrix.backend == 'llvm' && 'LLVM=1' || matrix.backend == 'cpu' && 'CPU=1' || matrix.backend == 'gpu' && 'GPU=1' }}" >> $GITHUB_ENV
+        run: printf "${{ matrix.backend == 'llvm' && 'LLVM=1' || matrix.backend == 'cpu' && 'CPU=1' || matrix.backend == 'gpu' && 'GPU=1' || matrix.backend == 'PTX' && 'FORWARD_ONLY=1\nJIT=1\nOPT=2\nCUDA=1\nPTX=1\nMOCKGPU=1' || matrix.backend == 'triton' && 'FORWARD_ONLY=1\nJIT=1\nOPT=2\nNV=1\nMOCKGPU=1\nTRITON=1\nTRITON_PTXAS_PATH=/usr/bin/ptxas' || matrix.backend == 'amd' && 'AMD=1\nMOCKGPU=1\nFORWARD_ONLY=1' || matrix.backend == 'nv' && 'NV=1\nMOCKGPU=1\nFORWARD_ONLY=1' }}" >> $GITHUB_ENV
      - name: Check Device.DEFAULT and print some source
        run: |
-          PYTHONPATH=${{ github.workspace }} python3 -c "from tinygrad import Device; assert Device.DEFAULT in ['LLVM','CPU','GPU'], Device.DEFAULT"
+          PYTHONPATH=${{ github.workspace }} python3 -c "from tinygrad import Device; assert Device.DEFAULT in ['LLVM','CPU','CUDA','GPU','AMD','NV'], Device.DEFAULT"
          DEBUG=5 PYTHONPATH=${{ github.workspace }} FORWARD_ONLY=1 python3 test/test_ops.py TestOps.test_add
-      - name: Run pytest (not cuda)
+      - name: Run pytest (not cuda or amd)
+        if: matrix.backend!='ptx' && matrix.backend!='triton' && matrix.backend != 'amd' && matrix.backend != 'nv'
        run: python -m pytest -n=auto test/ --ignore=test/models --ignore=test/unit --durations=20
+      - name: Run pytest (cuda)
+        if: matrix.backend=='ptx'||matrix.backend=='triton'||matrix.backend=='nv'
+        run: python -m pytest -n=auto test/ --ignore=test/models --ignore=test/unit --ignore test/test_gc.py --durations=20
+      - name: Run pytest (amd)
+        if: matrix.backend=='amd'
+        run: python -m pytest -n=auto test/test_ops.py test/test_dtype.py test/test_dtype_alu.py test/test_linearizer.py test/test_randomness.py test/imported/test_indexing.py test/test_hcq.py test/external/external_test_am.py --durations=20
      - name: Run TRANSCENDENTAL math
        run: TRANSCENDENTAL=2 python -m pytest -n=auto test/test_ops.py::TestOps::test_sin test/test_ops.py::TestOps::test_cos test/test_ops.py::TestOps::test_tan test/test_ops.py::TestOps::test_exp test/test_ops.py::TestOps::test_log --durations=20
      - name: Run process replay tests
@@ -677,8 +575,6 @@ jobs:
    name: MacOS (unit)
    runs-on: macos-14
    timeout-minutes: 20
-    env:
-      IGNORE_OOB: 0

    steps:
    - name: Checkout Code
@@ -691,7 +587,6 @@ jobs:
        python-version: '3.11'
        amd: 'true'
        cuda: 'true'
-        llvm: 'true'
    - name: Run real world test
      run: JIT=2 METAL=1 python -m pytest -n=auto test/models/test_real_world.py --durations=20
    - name: Test models (Metal)
@@ -719,14 +614,6 @@ jobs:
        FORWARD_ONLY: 1
      run: |
        python3 -m pytest -n=auto test/test_hcq.py test/test_tiny.py --durations=20
-    - name: Run pytest (amd with llvm backend)
-      env:
-        MOCKGPU: 1
-        AMD: 1
-        AMD_LLVM: 1
-        FORWARD_ONLY: 1
-      run: |
-        python -m pytest -n=auto test/test_hcq.py test/test_tiny.py test/test_amd_llvm.py --durations=20
    - name: Run pytest (ptx)
      env:
        MOCKGPU: 1
@@ -736,8 +623,6 @@ jobs:
      run: |
        python3 -m pytest -n=auto test/test_hcq.py test/test_tiny.py --durations=20
    - name: Run process replay tests
-      env:
-        IGNORE_OOB: 1
      uses: ./.github/actions/process-replay

  osxwebgpu:
@@ -769,8 +654,6 @@ jobs:
    name: MacOS (${{ matrix.backend }})
    runs-on: macos-15
    timeout-minutes: 20
-    env:
-      IGNORE_OOB: 0
    steps:
      - name: Checkout Code
        uses: actions/checkout@v4
@@ -804,9 +687,7 @@ jobs:

    name: Windows (${{ matrix.backend }})
    runs-on: windows-latest
-    timeout-minutes: 15
-    env:
-      IGNORE_OOB: 0
+    timeout-minutes: 10
    steps:
      - name: Checkout Code
        uses: actions/checkout@v4
@@ -11,7 +11,6 @@ notebooks
 *.txt
 build
 !examples/tinychat/assets/cdn.jsdelivr.net/npm/[email protected]/build/
-!examples/mlperf/training_submission_*/**/*.txt
 /dist
 *.egg-info
 /env
@@ -61,4 +60,3 @@ comgr_*
 site/
 profile_stats
 *.log
-target
@@ -78,11 +78,11 @@ generate_kfd() {
  clang2py /usr/include/linux/kfd_ioctl.h -o $BASE/kfd.py -k cdefstum

  fixup $BASE/kfd.py
-  sed -i "s/import ctypes/import ctypes, os/g" $BASE/kfd.py
-  sed -i "s/import fcntl, functools/import functools/g" $BASE/kfd.py
-  sed -i "/import functools/a from tinygrad.runtime.support.hcq import FileIOInterface" $BASE/kfd.py
-  sed -i "s/def _do_ioctl(__idir, __base, __nr, __user_struct, __fd, \*\*kwargs):/def _do_ioctl(__idir, __base, __nr, __user_struct, __fd:FileIOInterface, \*\*kwargs):/g" $BASE/kfd.py
-  sed -i "s/fcntl.ioctl(__fd, (__idir<<30)/__fd.ioctl((__idir<<30)/g" $BASE/kfd.py
+  sed -i "s\import ctypes\import ctypes, os\g" $BASE/kfd.py
+  sed -i "s\import fcntl, functools\import functools" $BASE/kfd.py
+  sed -i "s\import ctypes,os\a from tinygrad.runtime.support import HWInterface\g" $BASE/kfd.py
+  sed -i "s\def _do_ioctl(__idir, __base, __nr, __user_struct, __fd, **kwargs):\def _do_ioctl(__idir, __base, __nr, __user_struct, __fd:HWInterface, **kwargs):\g" $BASE/kfd.py
+  sed -i "s\fcntl.ioctl(__fd, (__idir<<30)\__fd.ioctl((__idir<<30)\g" $BASE/kfd.py
  python3 -c "import tinygrad.runtime.autogen.kfd"
 }

@@ -287,226 +287,79 @@ generate_vfio() {
  fixup $BASE/vfio.py
  sed -i "s\import ctypes\import ctypes, os\g" $BASE/vfio.py
  sed -i "s\import fcntl, functools\import functools" $BASE/vfio.py
-  sed -i "s\import ctypes,os\a from tinygrad.runtime.support import FileIOInterface\g" $BASE/vfio.py
+  sed -i "s\import ctypes,os\a from tinygrad.runtime.support import HWInterface\g" $BASE/vfio.py
  sed -i "s\fcntl.ioctl(__fd, (__idir<<30)\return __fd.ioctl((__idir<<30)\g" $BASE/vfio.py
 }

 generate_am() {
-  AMKERN_COMMIT_HASH=ceb12c04e2b5b53ec0779362831f5ee40c4921e4
-  AMKERN_SRC=/tmp/ROCK-Kernel-Driver-$AMKERN_COMMIT_HASH
-  if [ ! -d "$AMKERN_SRC" ]; then
-    git clone https://github.com/ROCm/ROCK-Kernel-Driver $AMKERN_SRC --depth 1
-  fi
-  AMKERN_AMD=$AMKERN_SRC/drivers/gpu/drm/amd/
-  AMKERN_INC=$AMKERN_AMD/include/
-
  clang2py -k cdefstum \
    extra/amdpci/headers/v11_structs.h \
-    extra/amdpci/headers/v12_structs.h \
    extra/amdpci/headers/amdgpu_vm.h \
    extra/amdpci/headers/discovery.h \
    extra/amdpci/headers/amdgpu_ucode.h \
+    extra/amdpci/headers/soc21_enum.h \
    extra/amdpci/headers/psp_gfx_if.h \
    extra/amdpci/headers/amdgpu_psp.h \
    extra/amdpci/headers/amdgpu_irq.h \
    extra/amdpci/headers/amdgpu_doorbell.h \
-    $AMKERN_INC/soc15_ih_clientid.h \
-    --clang-args="-include stdint.h" \
+    extra/amdpci/headers/soc15_ih_clientid.h \
    -o $BASE/am/am.py
  fixup $BASE/am/am.py
-  sed -i "s\(int64_t)\ \g" $BASE/am/am.py
-  sed -i "s\AMDGPU_PTE_MTYPE_VG10(2)\AMDGPU_PTE_MTYPE_VG10(0, 2)\g" $BASE/am/am.py # incorrect parsing (TODO: remove when clang2py is gone).

  clang2py -k cdefstum \
-    $AMKERN_AMD/amdkfd/kfd_pm4_headers_ai.h \
-    $AMKERN_AMD/amdgpu/soc15d.h \
-    -o $BASE/am/pm4_soc15.py
-  fixup $BASE/am/pm4_soc15.py
-
-  clang2py -k cdefstum \
-    $AMKERN_AMD/amdkfd/kfd_pm4_headers_ai.h \
-    $AMKERN_AMD/amdgpu/nvd.h \
-    -o $BASE/am/pm4_nv.py
-  fixup $BASE/am/pm4_nv.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/vega10_enum.h \
-    -o $BASE/am/vega10.py
-  fixup $BASE/am/vega10.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/navi10_enum.h \
-    -o $BASE/am/navi10.py
-  fixup $BASE/am/navi10.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/soc21_enum.h \
-    -o $BASE/am/soc21.py
-  fixup $BASE/am/soc21.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/soc24_enum.h \
-    -o $BASE/am/soc24.py
-  fixup $BASE/am/soc24.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mp/mp_13_0_0_offset.h \
-    $AMKERN_INC/asic_reg/mp/mp_13_0_0_sh_mask.h \
+    extra/amdpci/headers/mp_13_0_0_offset.h \
+    extra/amdpci/headers/mp_13_0_0_sh_mask.h \
    -o $BASE/am/mp_13_0_0.py
  fixup $BASE/am/mp_13_0_0.py

-  # 14_0_3 reuses 14_0_2
  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mp/mp_14_0_2_offset.h \
-    $AMKERN_INC/asic_reg/mp/mp_14_0_2_sh_mask.h \
-    -o $BASE/am/mp_14_0_3.py
-  fixup $BASE/am/mp_14_0_3.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mp/mp_11_0_offset.h \
-    $AMKERN_INC/asic_reg/mp/mp_11_0_sh_mask.h \
+    extra/amdpci/headers/mp_11_0_offset.h \
+    extra/amdpci/headers/mp_11_0_sh_mask.h \
    -o $BASE/am/mp_11_0.py
  fixup $BASE/am/mp_11_0.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/gc/gc_9_4_3_offset.h \
-    $AMKERN_INC/asic_reg/gc/gc_9_4_3_sh_mask.h \
-    extra/amdpci/overlay/gc_9_4_3.h \
-    -o $BASE/am/gc_9_4_3.py
-  fixup $BASE/am/gc_9_4_3.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/gc/gc_10_3_0_offset.h \
-    $AMKERN_INC/asic_reg/gc/gc_10_3_0_sh_mask.h \
-    -o $BASE/am/gc_10_3_0.py
-  fixup $BASE/am/gc_10_3_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/gc/gc_11_0_0_offset.h \
-    $AMKERN_INC/asic_reg/gc/gc_11_0_0_sh_mask.h \
+    extra/amdpci/headers/gc_11_0_0_offset.h \
+    extra/amdpci/headers/gc_11_0_0_sh_mask.h \
    -o $BASE/am/gc_11_0_0.py
  fixup $BASE/am/gc_11_0_0.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/gc/gc_12_0_0_offset.h \
-    $AMKERN_INC/asic_reg/gc/gc_12_0_0_sh_mask.h \
-    -o $BASE/am/gc_12_0_0.py
-  fixup $BASE/am/gc_12_0_0.py
-
-  clang2py -k cdefstum \
-    extra/hip_gpu_driver/sdma_registers.h \
-    $AMKERN_AMD/amdgpu/vega10_sdma_pkt_open.h \
-    --clang-args="-I/opt/rocm/include -x c++" \
-    -o $BASE/am/sdma_4_0_0.py
-  fixup $BASE/am/sdma_4_0_0.py
-
-  clang2py -k cdefstum \
-    extra/hip_gpu_driver/sdma_registers.h \
-    $AMKERN_AMD/amdgpu/navi10_sdma_pkt_open.h \
-    --clang-args="-I/opt/rocm/include -x c++" \
-    -o $BASE/am/sdma_5_0_0.py
-  fixup $BASE/am/sdma_5_0_0.py
-
-  clang2py -k cdefstum \
-    extra/hip_gpu_driver/sdma_registers.h \
-    $AMKERN_AMD/amdgpu/sdma_v6_0_0_pkt_open.h \
-    --clang-args="-I/opt/rocm/include -x c++" \
-    -o $BASE/am/sdma_6_0_0.py
-  fixup $BASE/am/sdma_6_0_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_3_0_0_offset.h \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_3_0_0_sh_mask.h \
+    extra/amdpci/headers/mmhub_3_0_0_offset.h \
+    extra/amdpci/headers/mmhub_3_0_0_sh_mask.h \
    -o $BASE/am/mmhub_3_0_0.py
  fixup $BASE/am/mmhub_3_0_0.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_3_0_2_offset.h \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_3_0_2_sh_mask.h \
+    extra/amdpci/headers/mmhub_3_0_2_offset.h \
+    extra/amdpci/headers/mmhub_3_0_2_sh_mask.h \
    -o $BASE/am/mmhub_3_0_2.py
  fixup $BASE/am/mmhub_3_0_2.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbio/nbio_2_3_offset.h \
-    $AMKERN_INC/asic_reg/nbio/nbio_2_3_sh_mask.h \
-    -o $BASE/am/nbio_2_3_0.py
-  fixup $BASE/am/nbio_2_3_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_2_0_offset.h \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_2_0_sh_mask.h \
-    -o $BASE/am/nbio_7_2_0.py
-  fixup $BASE/am/nbio_7_2_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_4_1_0_offset.h \
-    $AMKERN_INC/asic_reg/mmhub/mmhub_4_1_0_sh_mask.h \
-    -o $BASE/am/mmhub_4_1_0.py
-  fixup $BASE/am/mmhub_4_1_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbio/nbio_4_3_0_offset.h \
-    $AMKERN_INC/asic_reg/nbio/nbio_4_3_0_sh_mask.h \
+    extra/amdpci/headers/nbio_4_3_0_offset.h \
+    extra/amdpci/headers/nbio_4_3_0_sh_mask.h \
    -o $BASE/am/nbio_4_3_0.py
  fixup $BASE/am/nbio_4_3_0.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbif/nbif_6_3_1_offset.h \
-    $AMKERN_INC/asic_reg/nbif/nbif_6_3_1_sh_mask.h \
-    -o $BASE/am/nbif_6_3_1.py
-  fixup $BASE/am/nbif_6_3_1.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_9_0_offset.h \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_9_0_sh_mask.h \
-    -o $BASE/am/nbio_7_9_0.py
-  fixup $BASE/am/nbio_7_9_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_11_0_offset.h \
-    $AMKERN_INC/asic_reg/nbio/nbio_7_11_0_sh_mask.h \
-    -o $BASE/am/nbio_7_11_0.py
-  fixup $BASE/am/nbio_7_11_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/oss/osssys_6_0_0_offset.h \
-    $AMKERN_INC/asic_reg/oss/osssys_6_0_0_sh_mask.h \
+    extra/amdpci/headers/osssys_6_0_0_offset.h \
+    extra/amdpci/headers/osssys_6_0_0_sh_mask.h \
    -o $BASE/am/osssys_6_0_0.py
  fixup $BASE/am/osssys_6_0_0.py

  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/oss/osssys_7_0_0_offset.h \
-    $AMKERN_INC/asic_reg/oss/osssys_7_0_0_sh_mask.h \
-    -o $BASE/am/osssys_7_0_0.py
-  fixup $BASE/am/osssys_7_0_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_AMD/pm/swsmu/inc/pmfw_if/smu_v13_0_0_ppsmc.h \
-    $AMKERN_AMD/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_0.h \
+    extra/amdpci/headers/smu_v13_0_0_ppsmc.h \
+    extra/amdpci/headers/smu13_driver_if_v13_0_0.h \
    extra/amdpci/headers/amdgpu_smu.h \
    -o $BASE/am/smu_v13_0_0.py
  fixup $BASE/am/smu_v13_0_0.py

  clang2py -k cdefstum \
-    $AMKERN_AMD/pm/swsmu/inc/pmfw_if/smu_v14_0_0_pmfw.h \
-    $AMKERN_AMD/pm/swsmu/inc/pmfw_if/smu_v14_0_2_ppsmc.h \
-    $AMKERN_AMD/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h \
-    extra/amdpci/headers/amdgpu_smu.h \
-    --clang-args="-include stdint.h" \
-    -o $BASE/am/smu_v14_0_3.py
-  fixup $BASE/am/smu_v14_0_3.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/hdp/hdp_6_0_0_offset.h \
-    $AMKERN_INC/asic_reg/hdp/hdp_6_0_0_sh_mask.h \
+    extra/amdpci/headers/hdp_6_0_0_offset.h \
+    extra/amdpci/headers/hdp_6_0_0_sh_mask.h \
    -o $BASE/am/hdp_6_0_0.py
  fixup $BASE/am/hdp_6_0_0.py
-
-  clang2py -k cdefstum \
-    $AMKERN_INC/asic_reg/hdp/hdp_7_0_0_offset.h \
-    $AMKERN_INC/asic_reg/hdp/hdp_7_0_0_sh_mask.h \
-    -o $BASE/am/hdp_7_0_0.py
-  fixup $BASE/am/hdp_7_0_0.py
 }

 generate_sqtt() {
@@ -520,11 +373,10 @@ generate_sqtt() {
 }

 generate_webgpu() {
-  clang2py extra/webgpu/webgpu.h -o $BASE/webgpu.py
+  clang2py -l /usr/local/lib/libwebgpu_dawn.so extra/webgpu/webgpu.h -o $BASE/webgpu.py
  fixup $BASE/webgpu.py
-  sed -i "s/FIXME_STUB/webgpu/g" "$BASE/webgpu.py"
-  sed -i "s/FunctionFactoryStub()/ctypes.CDLL(webgpu_support.WEBGPU_PATH)/g" "$BASE/webgpu.py"
-  sed -i "s/import ctypes/import ctypes, tinygrad.runtime.support.webgpu as webgpu_support/g" "$BASE/webgpu.py"
+  sed -i 's/import ctypes/import ctypes, ctypes.util/g' $BASE/webgpu.py
+  sed -i "s|ctypes.CDLL('/usr/local/lib/libwebgpu_dawn.so')|ctypes.CDLL(ctypes.util.find_library('webgpu_dawn'))|g" $BASE/webgpu.py
  python3 -c "import tinygrad.runtime.autogen.webgpu"
 }

@@ -1,7 +1,7 @@
 # tinygrad is a tensor library, and as a tensor library it has multiple parts
 # 1. a "runtime". this allows buffer management, compilation, and running programs
 # 2. a "Device" that uses the runtime but specifies compute in an abstract way for all
-# 3. a "UOp" that fuses the compute into kernels, using memory only when needed
+# 3. a "LazyBuffer" that fuses the compute into kernels, using memory only when needed
 # 4. a "Tensor" that provides an easy to use frontend with autograd ".backward()"


@@ -74,47 +74,30 @@ fxn.exec([out, a, b])
 assert out.as_buffer().cast('I')[0] == 5


-print("******** third, the UOp ***********")
+print("******** third, the LazyBuffer ***********")

 from tinygrad.engine.realize import run_schedule
 from tinygrad.engine.schedule import create_schedule_with_vars
-from tinygrad.engine.grouper import get_becomes_map

 # allocate some values + load in values
-a = UOp.new_buffer(DEVICE, 1, dtypes.int32)
-b = UOp.new_buffer(DEVICE, 1, dtypes.int32)
+a = UOp.metaop(Ops.EMPTY, (1,), dtypes.int32, DEVICE)
+b = UOp.metaop(Ops.EMPTY, (1,), dtypes.int32, DEVICE)
 a.buffer.allocate().copyin(memoryview(bytearray(struct.pack("I", 2))))
 b.buffer.allocate().copyin(memoryview(bytearray(struct.pack("I", 3))))

 # describe the computation
-out = a + b
-s = UOp(Ops.SINK, dtypes.void, (out,))
+out = a.alu(Ops.ADD, b)

-# group the computation into kernels
-becomes_map = get_becomes_map(s)
-
-# the compute maps to an assign
-assign = becomes_map[a+b]
-
-# the first source is the output buffer (data)
-assert assign.src[0].op is Ops.BUFFER
-# the second source is the kernel (compute)
-assert assign.src[1].op is Ops.KERNEL
-
-# schedule the kernel graph in a linear list
-s = UOp(Ops.SINK, dtypes.void, (assign,))
-sched, _, becomes_map = create_schedule_with_vars(s)
-assert len(sched) == 1
+# schedule the computation as a list of kernels
+sched, _, becomes_map = create_schedule_with_vars(out.sink())
+for si in sched: print(si.ast.op)  # NOTE: the first two convert it to CPU
+# NOTE: UOps are no longer mutable, the scheduler gives you a map to lookup which BUFFER the result was written to
+out = becomes_map[out]

 # DEBUGGING: print the compute ast
 print(sched[-1].ast)
 # NOTE: sched[-1].ast is the same as st_0 above

-# the output will be stored in a new buffer
-out = becomes_map[assign]
-assert out.op is Ops.BUFFER and not out.buffer.is_allocated()
-print(out)
-
 # run that schedule
 run_schedule(sched)

@@ -1,6 +1,6 @@
 # AM Driver

-AM driver is a userspace driver targeting AMD's RDNA3/RDNA4. You only need tinygrad to send compute tasks to your GPU!
+AM driver is a userspace driver targeting AMD's 7900XTX. You only need tinygrad to send compute tasks to your GPU!

 ## How to run?
 Make sure that amdgpu module is unloaded and just run tinygrad with `AMD=1`!
@@ -115,8 +115,9 @@ HCQ-compatible devices use a global timeline signal for synchronizing all operat
 ```python
 HWQueue().wait(your_device.timeline_signal, your_device.timeline_value - 1) \
         .exec(...)
-         .signal(your_device.timeline_signal, your_device.next_timeline()) \
+         .signal(your_device.timeline_signal, your_device.timeline_value) \
         .submit(your_device)
+your_device.timeline_value += 1

 # Optionally wait for execution
 your_device.timeline_signal.wait(your_device.timeline_value - 1)
@@ -1,109 +0,0 @@
-# Kernel Creation
-
-Tinygrad lazily builds up a graph of Tensor operations. The Tensor graph includes a mix of:
-
- Buffer and Assignment Ops: `BUFFER`, `BUFFER_VIEW`, `COPY`, `ASSIGN`
- Movement Ops: `RESHAPE`, `EXPAND`, `PERMUTE`, `PAD`, `SHRINK`, `FLIP`
- Compute Ops: `ADD`, `MUL`, `REDUCE_AXIS`, ...
-
-`Tensor.kernelize` creates the kernels and buffers needed to realize the output Tensor(s).
-
-## Kernelize flow
-
-Let's see how a multiply add Tensor graph becomes a fused elementwise kernel.
-
-```py
-# initialize 3 input buffers on the device
-a = Tensor([1]).realize()
-b = Tensor([2]).realize()
-c = Tensor([3]).realize()
-
-# create the Tensor graph
-mul = a*b
-out = mul+c
-
-print(mul) # <Tensor <UOp METAL (1,) int (<Ops.MUL: 48>, None)> on METAL with grad None>
-print(out) # <Tensor <UOp METAL (1,) int (<Ops.ADD: 52>, None)> on METAL with grad None>
-
-out.kernelize()
-
-print(mul) # <Tensor <UOp METAL (1,) int (<Ops.MUL: 48>, None)> on METAL with grad None>
-print(out) # <Tensor <UOp METAL (1,) int (<Ops.ASSIGN: 66>, None)> on METAL with grad None>
-```
-
-The multiply Tensor stays the same because it is fused. The output Tensor's UOp becomes a new ASSIGN UOp:
-
-```py
-print(out.lazydata)
-```
-
-The first source is the output BUFFER:
-
-```
-UOp(Ops.BUFFER, dtypes.int, arg=1, src=(
-  UOp(Ops.DEVICE, dtypes.void, arg='METAL', src=()),
-  UOp(Ops.UNIQUE, dtypes.void, arg=6, src=()),))
-```
-
-And the second source is the KERNEL and its 4 buffer edges (output_buffer, a, b, c):
-
-```
-UOp(Ops.KERNEL, dtypes.void, arg=<Kernel 12 SINK(<Ops.STORE: 45>,) (__add__, __mul__)>, src=(
-  UOp(Ops.BUFFER, dtypes.int, arg=1, src=(
-    x1:=UOp(Ops.DEVICE, dtypes.void, arg='METAL', src=()),
-    UOp(Ops.UNIQUE, dtypes.void, arg=6, src=()),)),
-  UOp(Ops.BUFFER, dtypes.int, arg=1, src=(
-     x1,
-    UOp(Ops.UNIQUE, dtypes.void, arg=1, src=()),)),
-  UOp(Ops.BUFFER, dtypes.int, arg=1, src=(
-     x1,
-    UOp(Ops.UNIQUE, dtypes.void, arg=3, src=()),)),
-  UOp(Ops.BUFFER, dtypes.int, arg=1, src=(
-     x1,
-    UOp(Ops.UNIQUE, dtypes.void, arg=5, src=()),)),))
-```
-
-KERNEL describes the compute AST, metadata and memory dependencies.
-
-BUFFER holds a reference to the device memory where the output will be stored.
-
-Once a Tensor is kernelized, all children will LOAD its BUFFER, instead of fusing it:
-
-```py
-child = out+2
-child.kernelize()
-print(child.lazydata.src[1].arg.ast)
-```
-
-```
-UOp(Ops.SINK, dtypes.void, arg=None, src=(
-  UOp(Ops.STORE, dtypes.void, arg=None, src=(
-    UOp(Ops.DEFINE_GLOBAL, dtypes.int.ptr(1), arg=0, src=()),
-    x2:=UOp(Ops.VIEW, dtypes.void, arg=ShapeTracker(views=(View(shape=(1,), strides=(0,), offset=0, mask=None, contiguous=True),)), src=()),
-    UOp(Ops.ADD, dtypes.int, arg=None, src=(
-      UOp(Ops.LOAD, dtypes.int, arg=None, src=(
-        UOp(Ops.DEFINE_GLOBAL, dtypes.int.ptr(1), arg=1, src=()),
-         x2,)),
-      UOp(Ops.CONST, dtypes.int, arg=2, src=(
-         x2,)),)),)),))
-```
-
-`Tensor.realize` will execute the kernels and write outputs to memory:
-
-```py
-Tensor.realize(out)
-print(out)        # <Tensor <UOp METAL (1,) int (<Ops.BUFFER: 23>, <buf real:True device:METAL size:1 dtype:dtypes.int offset:0>)> on METAL with grad None>
-print(out.item()) # 5
-```
-
-<hr />
-
-**Summary**
-
- The large Tensor graph is built from a mix of data, compute and movement Ops.
-
- `Tensor.kernelize` splits the Tensor graph into data (BUFFER), compute (KERNEL) and links dependencies with ASSIGN.
-
- `Tensor.realize` executes KERNELs on device and replaces the Tensor graph with just a BUFFER.
-
- Kernelize can be called multiple times on a Tensor. This allows for incrementally building the kernel fusion layout of a large Tensor graph, without having to call `realize` or `schedule`.
@@ -30,7 +30,7 @@ These control the behavior of core tinygrad even when used as a library.

 Variable | Possible Value(s) | Description
 ---|---|---
-DEBUG               | [1-7]      | enable debugging output (operations, timings, speed, generated code and more)
+DEBUG               | [1-6]      | enable debugging output, with 4 you get operations, timings, speed, generated code and more
 GPU                 | [1]        | enable the GPU (OpenCL) backend
 CUDA                | [1]        | enable CUDA backend
 AMD                 | [1]        | enable AMD backend
@@ -49,17 +49,4 @@ VISIBLE_DEVICES     | [list[int]]| restricts the NV/AMD devices that are availab
 JIT                 | [0-2]      | 0=disabled, 1=[jit enabled](quickstart.md#jit) (default), 2=jit enabled, but graphs are disabled
 VIZ                 | [1]        | 0=disabled, 1=[viz enabled](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/viz)
 ALLOW_TF32          | [1]        | enable TensorFloat-32 tensor cores on Ampere or newer GPUs.
-WEBGPU_BACKEND      | [WGPUBackendType_Metal, ...]          | Force select a backend for WebGPU (Metal, DirectX, OpenGL, Vulkan...)
-CUDA_PATH           | str        | Use `CUDA_PATH/include` for CUDA headers for CUDA and NV backends. If not set, TinyGrad will use `/usr/local/cuda/include`, `/usr/include` and `/opt/cuda/include`.
-
-## Debug breakdown
-
-Variable | Value | Description
---|---|---
-DEBUG               | >= 1       | Enables debugging and lists devices being used
-DEBUG               | >= 2       | Provides performance metrics for operations, including timing, memory usage, bandwidth for each kernel execution
-DEBUG               | >= 3       | Outputs buffers used for each kernel (shape, dtype and strides) and the applied optimizations at a kernel level
-DEBUG               | >= 4       | Outputs the generated kernel code
-DEBUG               | >= 5       | Displays the intermediate representation of the computation UOps (AST)
-DEBUG               | >= 6       | Displays the intermediate representation of the computation UOps in a linearized manner, detailing the operation sequence
-DEBUG               | >= 7       | Outputs the assembly code generated for the target hardware
+WEBGPU_BACKEND      | [WGPUBackendType_Metal, ...]          | Force select a backend for WebGPU (Metal, DirectX, OpenGL, Vulkan...)
@@ -29,12 +29,5 @@
 ::: tinygrad.nn.state.get_state_dict
 ::: tinygrad.nn.state.get_parameters
 ::: tinygrad.nn.state.load_state_dict
-::: tinygrad.nn.state.tar_extract
-    options:
-        show_signature: false
-        separate_signature: false
 ::: tinygrad.nn.state.torch_load
-    options:
-        show_signature: false
-        separate_signature: false
 ::: tinygrad.nn.state.gguf_load
@@ -5,7 +5,7 @@ tinygrad supports various runtimes, enabling your code to scale across a wide ra
 | Runtime | Description | Requirements |
 |---------|-------------|--------------|
 | [NV](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_nv.py) | Provides acceleration for NVIDIA GPUs | Ampere/Ada series GPUs |
-| [AMD](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_amd.py) | Provides acceleration for AMD GPUs | RDNA2/RDNA3/RDNA4 series GPUs |
+| [AMD](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_amd.py) | Provides acceleration for AMD GPUs | RDNA2/RDNA3 series GPUs |
 | [QCOM](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_qcom.py) | Provides acceleration for QCOM GPUs | 6xx series GPUs |
 | [METAL](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_metal.py) | Utilizes Metal for acceleration on Apple devices | M1+ Macs; Metal 3.0+ for `bfloat` support |
 | [CUDA](https://github.com/tinygrad/tinygrad/tree/master/tinygrad/runtime/ops_cuda.py) | Utilizes CUDA for acceleration on NVIDIA GPUs | NVIDIA GPU with CUDA support |
@@ -20,7 +20,6 @@

 ::: tinygrad.Tensor.manual_seed
 ::: tinygrad.Tensor.rand
-::: tinygrad.Tensor.rand_like
 ::: tinygrad.Tensor.randn
 ::: tinygrad.Tensor.randint
 ::: tinygrad.Tensor.normal
@@ -22,7 +22,6 @@ Elementwise ops operate on a per element basis. They don't change the shape of t
 ::: tinygrad.Tensor.round
 ::: tinygrad.Tensor.isinf
 ::: tinygrad.Tensor.isnan
-::: tinygrad.Tensor.isfinite
 ::: tinygrad.Tensor.lerp
 ::: tinygrad.Tensor.square
 ::: tinygrad.Tensor.clamp
@@ -69,14 +68,12 @@ Elementwise ops operate on a per element basis. They don't change the shape of t
 ::: tinygrad.Tensor.bitwise_xor
 ::: tinygrad.Tensor.bitwise_and
 ::: tinygrad.Tensor.bitwise_or
-::: tinygrad.Tensor.bitwise_not
 ::: tinygrad.Tensor.lshift
 ::: tinygrad.Tensor.rshift
 ::: tinygrad.Tensor.pow
 ::: tinygrad.Tensor.maximum
 ::: tinygrad.Tensor.minimum
 ::: tinygrad.Tensor.where
-::: tinygrad.Tensor.copysign

 ## Casting Ops

@@ -9,7 +9,6 @@
 ::: tinygrad.Tensor.isclose
 ::: tinygrad.Tensor.mean
 ::: tinygrad.Tensor.var
-::: tinygrad.Tensor.var_mean
 ::: tinygrad.Tensor.std
 ::: tinygrad.Tensor.std_mean
 ::: tinygrad.Tensor.softmax
@@ -23,7 +22,6 @@

 ::: tinygrad.Tensor.avg_pool2d
 ::: tinygrad.Tensor.max_pool2d
-::: tinygrad.Tensor.max_unpool2d
 ::: tinygrad.Tensor.conv2d
 ::: tinygrad.Tensor.conv_transpose2d
 ::: tinygrad.Tensor.dot
@@ -36,7 +34,6 @@
 ::: tinygrad.Tensor.interpolate
 ::: tinygrad.Tensor.scatter
 ::: tinygrad.Tensor.scatter_reduce
-::: tinygrad.Tensor.masked_select
 ::: tinygrad.Tensor.sort
 ::: tinygrad.Tensor.topk

@@ -25,7 +25,6 @@
 ::: tinygrad.Tensor.replace
 ::: tinygrad.Tensor.assign
 ::: tinygrad.Tensor.detach
-::: tinygrad.Tensor.clone
 ::: tinygrad.Tensor.to
 ::: tinygrad.Tensor.to_
 ::: tinygrad.Tensor.shard
@@ -1,12 +1,12 @@
-import sys, onnx, time, pickle
-from tinygrad import TinyJit, GlobalCounters, fetch, getenv
-from tinygrad.frontend.onnx import OnnxRunner
+import sys, onnx, time
+from tinygrad import TinyJit, Device, GlobalCounters, fetch, getenv
+from extra.onnx import OnnxRunner
 from extra.onnx_helpers import get_example_inputs, validate

 def load_onnx_model(onnx_file):
  onnx_model = onnx.load(onnx_file)
  run_onnx = OnnxRunner(onnx_model)
-  run_onnx_jit = TinyJit(lambda **kwargs: next(iter(run_onnx({k:v.to(None) for k,v in kwargs.items()}).values())), prune=True, optimize=True)
+  run_onnx_jit = TinyJit(lambda **kwargs: next(iter(run_onnx({k:v.to(Device.DEFAULT) for k,v in kwargs.items()}).values())), prune=True)
  return run_onnx_jit, run_onnx.graph_inputs

 if __name__ == "__main__":
@@ -33,4 +33,4 @@ if __name__ == "__main__":

  if getenv("ORT"):
    validate(onnx_file, new_inputs, rtol=1e-3, atol=1e-3)
-    print("model validated")
+    print("model validated")
@@ -8,7 +8,7 @@ import numpy as np
 import subprocess
 import tensorflow as tf
 import tf2onnx
-from tinygrad.frontend.onnx import OnnxRunner
+from extra.onnx import OnnxRunner
 from tinygrad.tensor import Tensor
 from extra.export_model import export_model_clang, compile_net, jit_model

@@ -189,7 +189,7 @@ class GPT2:
          tokens = Variable("tokens", 0, VOCAB_SIZE).bind(toks[0][start_pos])
        else:
          tokens = Tensor([x[start_pos:] for x in toks])
-        tok = self.model(tokens, Variable("start_pos", 1 if start_pos else 0, MAX_CONTEXT-1).bind(start_pos), temperature).tolist()
+        tok = self.model(tokens, Variable("start_pos", 1 if start_pos else 0, MAX_CONTEXT).bind(start_pos), temperature).tolist()
      start_pos = len(toks[0])
      for i,t in enumerate(tok): toks[i].append(t)
    return [self.tokenizer.decode(x) for x in toks]
@@ -3,7 +3,6 @@ from extra.mcts_search import mcts_search
 from examples.mlperf.helpers import get_mlperf_bert_model
 from tinygrad import Tensor, Device, dtypes, nn
 from tinygrad.codegen.kernel import Kernel
-from tinygrad.codegen.heuristic import hand_coded_optimizations
 from tinygrad.ops import Ops, sym_infer
 from tinygrad.device import Compiled
 from tinygrad.engine.search import beam_search, bufs_from_lin
@@ -84,7 +83,7 @@ if __name__ == "__main__":

    # always try hand coded opt
    lin = Kernel(si.ast, opts=device.renderer)
-    lin.apply_opts(hand_coded_optimizations(lin))
+    lin.hand_coded_optimizations()
    lins.append((lin, "HC"))

    # maybe try tensor cores
@@ -348,167 +348,6 @@ def batch_load_unet3d(preprocessed_dataset_dir:Path, batch_size:int=6, val:bool=
      # happens with BENCHMARK set
      pass

-### RetinaNet
-
-def load_retinanet_data(base_dir:Path, val:bool, queue_in:Queue, queue_out:Queue,
-                        imgs:Tensor, boxes:Tensor, labels:Tensor, matches:Tensor|None=None,
-                        anchors:Tensor|None=None, seed:int|None=None):
-  from extra.datasets.openimages import image_load, random_horizontal_flip, resize
-  from examples.mlperf.helpers import box_iou, find_matches, generate_anchors
-  import torch
-
-  while (data:=queue_in.get()) is not None:
-    idx, img, tgt = data
-    img = image_load(base_dir, img["subset"], img["file_name"])
-
-    if val:
-      img = resize(img)[0]
-    else:
-      if seed is not None:
-        np.random.seed(seed)
-        random.seed(seed)
-        torch.manual_seed(seed)
-
-      img, tgt = random_horizontal_flip(img, tgt)
-      img, tgt, _ = resize(img, tgt=tgt)
-      match_quality_matrix = box_iou(tgt["boxes"], (anchor := np.concatenate(generate_anchors((800, 800)))))
-      match_idxs = find_matches(match_quality_matrix, allow_low_quality_matches=True)
-      clipped_match_idxs = np.clip(match_idxs, 0, None)
-      clipped_boxes, clipped_labels = tgt["boxes"][clipped_match_idxs], tgt["labels"][clipped_match_idxs]
-
-      boxes[idx].contiguous().realize().lazydata.base.realized.as_buffer(force_zero_copy=True)[:] = clipped_boxes.tobytes()
-      labels[idx].contiguous().realize().lazydata.base.realized.as_buffer(force_zero_copy=True)[:] = clipped_labels.tobytes()
-      matches[idx].contiguous().realize().lazydata.base.realized.as_buffer(force_zero_copy=True)[:] = match_idxs.tobytes()
-      anchors[idx].contiguous().realize().lazydata.base.realized.as_buffer(force_zero_copy=True)[:] = anchor.tobytes()
-
-    imgs[idx].contiguous().realize().lazydata.base.realized.as_buffer(force_zero_copy=True)[:] = img.tobytes()
-
-    queue_out.put(idx)
-  queue_out.put(None)
-
-def batch_load_retinanet(dataset, val:bool, base_dir:Path, batch_size:int=32, shuffle:bool=True, seed:int|None=None):
-  def _enqueue_batch(bc):
-    from extra.datasets.openimages import prepare_target
-    for idx in range(bc * batch_size, (bc+1) * batch_size):
-      img = dataset.loadImgs(next(dataset_iter))[0]
-      ann = dataset.loadAnns(dataset.getAnnIds(img_id:=img["id"]))
-      tgt = prepare_target(ann, img_id, (img["height"], img["width"]))
-
-      if img_ids is not None:
-        img_ids[idx] = img_id
-
-      if img_sizes is not None:
-        img_sizes[idx] = tgt["image_size"]
-
-      queue_in.put((idx, img, tgt))
-
-  def _setup_shared_mem(shm_name:str, size:tuple[int, ...], dtype:dtypes) -> tuple[shared_memory.SharedMemory, Tensor]:
-    if os.path.exists(f"/dev/shm/{shm_name}"): os.unlink(f"/dev/shm/{shm_name}")
-    shm = shared_memory.SharedMemory(name=shm_name, create=True, size=prod(size))
-    shm_tensor = Tensor.empty(*size, dtype=dtype, device=f"disk:/dev/shm/{shm_name}")
-    return shm, shm_tensor
-
-  image_ids = sorted(dataset.imgs.keys())
-  batch_count = min(32, len(image_ids) // batch_size)
-
-  queue_in, queue_out = Queue(), Queue()
-  procs, data_out_count = [], [0] * batch_count
-
-  shm_imgs, imgs = _setup_shared_mem("retinanet_imgs", (batch_size * batch_count, 800, 800, 3), dtypes.uint8)
-
-  if val:
-    boxes, labels, matches, anchors = None, None, None, None
-    img_ids, img_sizes = [None] * (batch_size * batch_count), [None] * (batch_size * batch_count)
-  else:
-    img_ids, img_sizes = None, None
-    shm_boxes, boxes = _setup_shared_mem("retinanet_boxes", (batch_size * batch_count, 120087, 4), dtypes.float32)
-    shm_labels, labels = _setup_shared_mem("retinanet_labels", (batch_size * batch_count, 120087), dtypes.int64)
-    shm_matches, matches = _setup_shared_mem("retinanet_matches", (batch_size * batch_count, 120087), dtypes.int64)
-    shm_anchors, anchors = _setup_shared_mem("retinanet_anchors", (batch_size * batch_count, 120087, 4), dtypes.float64)
-
-  shutdown = False
-  class Cookie:
-    def __init__(self, bc):
-      self.bc = bc
-    def __del__(self):
-      if not shutdown:
-        try: _enqueue_batch(self.bc)
-        except StopIteration: pass
-
-  def shuffle_indices(indices, seed):
-    rng = random.Random(seed)
-    rng.shuffle(indices)
-
-  if shuffle: shuffle_indices(image_ids, seed=seed)
-  dataset_iter = iter(image_ids)
-
-  try:
-    for _ in range(cpu_count()):
-      proc = Process(
-        target=load_retinanet_data,
-        args=(base_dir, val, queue_in, queue_out, imgs, boxes, labels),
-        kwargs={"matches": matches, "anchors": anchors, "seed": seed}
-      )
-      proc.daemon = True
-      proc.start()
-      procs.append(proc)
-
-    for bc in range(batch_count):
-      _enqueue_batch(bc)
-
-    for _ in range(len(image_ids) // batch_size):
-      while True:
-        bc = queue_out.get() // batch_size
-        data_out_count[bc] += 1
-        if data_out_count[bc] == batch_size: break
-
-      data_out_count[bc] = 0
-
-      if val:
-        yield (imgs[bc * batch_size:(bc + 1) * batch_size],
-               img_ids[bc * batch_size:(bc + 1) * batch_size],
-               img_sizes[bc * batch_size:(bc + 1) * batch_size],
-               Cookie(bc))
-      else:
-        yield (imgs[bc * batch_size:(bc + 1) * batch_size],
-               boxes[bc * batch_size:(bc + 1) * batch_size],
-               labels[bc * batch_size:(bc + 1) * batch_size],
-               matches[bc * batch_size:(bc + 1) * batch_size],
-               anchors[bc * batch_size:(bc + 1) * batch_size],
-               Cookie(bc))
-  finally:
-    shutdown = True
-
-    for _ in procs: queue_in.put(None)
-    queue_in.close()
-
-    for _ in procs:
-      while queue_out.get() is not None: pass
-    queue_out.close()
-
-    # shutdown processes
-    for proc in procs: proc.join()
-
-    shm_imgs.close()
-
-    if not val:
-      shm_boxes.close()
-      shm_labels.close()
-      shm_matches.close()
-      shm_anchors.close()
-
-    try:
-      shm_imgs.unlink()
-
-      if not val:
-        shm_boxes.unlink()
-        shm_labels.unlink()
-        shm_matches.unlink()
-        shm_anchors.unlink()
-    except FileNotFoundError:
-      # happens with BENCHMARK set
-      pass
-
 if __name__ == "__main__":
  def load_unet3d(val):
    assert not val, "validation set is not supported due to different sizes on inputs"
@@ -529,14 +368,6 @@ if __name__ == "__main__":
      for x,y,c in batch_load_resnet(val=val):
        pbar.update(x.shape[0])

-  def load_retinanet(val):
-    from extra.datasets.openimages import BASEDIR, download_dataset
-    from pycocotools.coco import COCO
-    dataset = COCO(download_dataset(base_dir:=getenv("BASE_DIR", BASEDIR), "validation" if val else "train"))
-    with tqdm(total=len(dataset.imgs.keys())) as pbar:
-      for x in batch_load_retinanet(dataset, val, base_dir):
-        pbar.update(x[0].shape[0])
-
  load_fn_name = f"load_{getenv('MODEL', 'resnet')}"
  if load_fn_name in globals():
    globals()[load_fn_name](getenv("VAL", 1))
@@ -1,7 +1,6 @@
 from collections import OrderedDict
 import unicodedata
 from typing import Optional
-import math
 import numpy as np
 from tinygrad.nn import state
 from tinygrad.tensor import Tensor, dtypes
@@ -231,126 +230,3 @@ def get_fake_data_bert(BS:int):
    "masked_lm_weights": Tensor.empty((BS, 76), dtype=dtypes.float32, device="CPU"),
    "next_sentence_labels": Tensor.empty((BS, 1), dtype=dtypes.int32, device="CPU"),
  }
-
-def find_matches(match_quality_matrix:np.ndarray, high_threshold:float=0.5, low_threshold:float=0.4, allow_low_quality_matches:bool=False) -> np.ndarray:
-  BELOW_LOW_THRESHOLD, BETWEEN_THRESHOLDS = -1, -2
-
-  def _set_low_quality_matches_(matches:np.ndarray, all_matches:np.ndarray, match_quality_matrix:np.ndarray):
-    highest_quality_foreach_gt = np.max(match_quality_matrix, axis=1)
-    pred_inds_to_update = np.nonzero(match_quality_matrix == highest_quality_foreach_gt[:, None])[1]
-    matches[pred_inds_to_update] = all_matches[pred_inds_to_update]
-
-  assert low_threshold <= high_threshold
-
-  matched_vals, matches = match_quality_matrix.max(axis=0), match_quality_matrix.argmax(axis=0)
-  all_matches = np.copy(matches) if allow_low_quality_matches else None
-  below_low_threshold = matched_vals < low_threshold
-  between_thresholds = (matched_vals >= low_threshold) & (matched_vals < high_threshold)
-  matches[below_low_threshold] = BELOW_LOW_THRESHOLD
-  matches[between_thresholds] = BETWEEN_THRESHOLDS
-
-  if allow_low_quality_matches:
-    assert all_matches is not None
-    _set_low_quality_matches_(matches, all_matches, match_quality_matrix)
-
-  return matches
-
-def box_iou(boxes1:np.ndarray, boxes2:np.ndarray) -> np.ndarray:
-  def _box_area(boxes:np.ndarray) -> np.ndarray: return (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
-
-  def _box_inter_union(boxes1:np.ndarray, boxes2:np.ndarray) -> tuple[np.ndarray, np.ndarray]:
-    area1, area2 = _box_area(boxes1), _box_area(boxes2)
-    lt, rb = np.maximum(boxes1[:, None, :2], boxes2[:, :2]), np.minimum(boxes1[:, None, 2:], boxes2[:, 2:])
-    wh = np.clip(rb - lt, a_min=0, a_max=None)
-    inter = wh[:, :, 0] * wh[:, :, 1]
-    union = area1[:, None] + area2 - inter
-    return inter, union
-
-  inter, union = _box_inter_union(boxes1, boxes2)
-  return inter / union
-
-def generate_anchors(input_size:tuple[int, int], scales:Optional[tuple[Tensor, ...]]=None, aspect_ratios:Optional[tuple[Tensor, ...]]=None) -> list[np.ndarray]:
-  def _compute_grid_sizes(input_size:tuple[int, int]) -> np.ndarray:
-    return np.ceil(np.array(input_size)[None, :] / 2 ** np.arange(3, 8)[:, None])
-
-  scales = tuple((i, int(i * 2 ** (1/3)), int(i * 2 ** (2/3))) for i in 2 ** np.arange(5, 10)) if scales is None else scales
-  aspect_ratios = ((0.5, 1.0, 2.0),) * len(scales) if aspect_ratios is None else aspect_ratios
-  aspect_ratios = tuple(ar for ar in aspect_ratios)
-  grid_sizes = _compute_grid_sizes(input_size)
-
-  assert len(scales) == len(aspect_ratios) == len(grid_sizes), "scales, aspect_ratios, and grid_sizes must have the same length"
-
-  anchors = []
-  for s, ar, gs in zip(scales, aspect_ratios, grid_sizes):
-    s, ar = np.array(s), np.array(ar)
-    h_ratios = np.sqrt(ar)
-    w_ratios = 1 / h_ratios
-    ws = (w_ratios[:, None] * s[None, :]).reshape(-1)
-    hs = (h_ratios[:, None] * s[None, :]).reshape(-1)
-    base_anchors = (np.stack([-ws, -hs, ws, hs], axis=1) / 2).round()
-    stride_h, stride_w = input_size[0] // gs[0], input_size[1] // gs[1]
-    shifts_x, shifts_y = np.meshgrid(np.arange(gs[1]) * stride_w, np.arange(gs[0]) * stride_h)
-    shifts_x, shifts_y = shifts_x.reshape(-1), shifts_y.reshape(-1)
-    shifts = np.stack([shifts_x, shifts_y, shifts_x, shifts_y], axis=1, dtype=np.float32)
-    anchors.append((shifts[:, None] + base_anchors[None, :]).reshape(-1, 4))
-
-  return anchors
-
-
-class BoxCoder(object):
-  def __init__(self, weights, bbox_xform_clip=math.log(1000. / 16), apply_to_remove=True):
-    self.weights = weights
-    self.bbox_xform_clip = bbox_xform_clip
-    self.apply_to_remove = apply_to_remove
-
-  def encode(self, reference_boxes, proposals):
-    TO_REMOVE = self.apply_to_remove  # TODO remove
-    ex_widths = proposals[..., 2] - proposals[..., 0] + TO_REMOVE
-    ex_heights = proposals[..., 3] - proposals[..., 1] + TO_REMOVE
-    ex_ctr_x = proposals[..., 0] + 0.5 * ex_widths
-    ex_ctr_y = proposals[..., 1] + 0.5 * ex_heights
-
-    gt_widths = reference_boxes[..., 2] - reference_boxes[..., 0] + TO_REMOVE
-    gt_heights = reference_boxes[..., 3] - reference_boxes[..., 1] + TO_REMOVE
-    gt_ctr_x = reference_boxes[..., 0] + 0.5 * gt_widths
-    gt_ctr_y = reference_boxes[..., 1] + 0.5 * gt_heights
-
-    wx, wy, ww, wh = self.weights
-    targets_dx = wx * (gt_ctr_x - ex_ctr_x) / ex_widths
-    targets_dy = wy * (gt_ctr_y - ex_ctr_y) / ex_heights
-    targets_dw = ww * Tensor.log(gt_widths / ex_widths)
-    targets_dh = wh * Tensor.log(gt_heights / ex_heights)
-
-    targets = Tensor.stack(targets_dx, targets_dy, targets_dw, targets_dh, dim=-1)
-    return targets
-
-  def decode(self, rel_codes, boxes):
-    boxes = boxes.cast(rel_codes.dtype)
-    rel_codes = rel_codes
-
-    TO_REMOVE = self.apply_to_remove  # TODO remove
-    widths = boxes[:, 2] - boxes[:, 0] + TO_REMOVE
-    heights = boxes[:, 3] - boxes[:, 1] + TO_REMOVE
-    ctr_x = boxes[:, 0] + 0.5 * widths
-    ctr_y = boxes[:, 1] + 0.5 * heights
-
-    wx, wy, ww, wh = self.weights
-    dx = rel_codes[:, 0::4] / wx
-    dy = rel_codes[:, 1::4] / wy
-    dw = rel_codes[:, 2::4] / ww
-    dh = rel_codes[:, 3::4] / wh
-
-    # Prevent sending too large values into Tensor.exp()
-    dw = dw.clip(min_=dw.min(), max_=self.bbox_xform_clip)
-    dh = dh.clip(min_=dh.min(), max_=self.bbox_xform_clip)
-
-    pred_ctr_x = dx * widths[:, None] + ctr_x[:, None]
-    pred_ctr_y = dy * heights[:, None] + ctr_y[:, None]
-    pred_w = dw.exp() * widths[:, None]
-    pred_h = dh.exp() * heights[:, None]
-    x = pred_ctr_x - 0.5 * pred_w
-    y = pred_ctr_y - 0.5 * pred_h
-    w = pred_ctr_x + 0.5 * pred_w - 1
-    h = pred_ctr_y + 0.5 * pred_h - 1
-    pred_boxes = Tensor.stack(x, y, w, h).permute(1,2,0).reshape(rel_codes.shape[0], rel_codes.shape[1])
-    return pred_boxes
@@ -53,9 +53,7 @@ class EmbeddingBert(nn.Embedding):
    arange_shp, weight_shp, big_shp = (1, 1, self.vocab_sz, 1), (1, 1, self.vocab_sz, self.embed_sz), idx.shape+(self.vocab_sz, self.embed_sz,)
    if not hasattr(self, 'arange'): self.arange = Tensor.arange(self.vocab_sz, requires_grad=False, device=self.weight.device).reshape(arange_shp)
    arange, idx, vals = self.arange.expand(big_shp), idx.reshape(idx.shape+(1, 1,)).expand(big_shp), self.weight.cast(dtypes.default_float).reshape(weight_shp).expand(big_shp)
-    # TODO: contiguous() here because the embedding dropout creates different asts on each device, and search becomes very slow.
-    # Should fix with fixing random ast on multi device, and fuse arange to make embedding fast.
-    return (arange == idx).mul(vals).sum(2, dtype=vals.dtype).contiguous()
+    return (arange == idx).mul(vals).sum(2, dtype=vals.dtype)

 class LayerNormBert:
  def __init__(self, normalized_shape:Union[int, tuple[int, ...]], eps:float=1e-12, elementwise_affine:bool=True):
@@ -68,62 +66,3 @@ class LayerNormBert:
    xn = x.cast(dtypes.float32).layernorm(eps=self.eps, axis=self.axis).cast(x.dtype)
    if not self.elementwise_affine: return xn
    return (xn * self.weight.cast(dtypes.default_float) + self.bias.cast(dtypes.default_float))
-
-class FrozenBatchNorm2dRetinaNet(nn.BatchNorm2d):
-  def __init__(self, sz:int, eps=1e-5, affine=True, track_running_stats=True, momentum=0.1):
-    self.eps, self.track_running_stats, self.momentum = eps, track_running_stats, momentum
-
-    self.weight = Tensor.ones(sz, dtype=dtypes.float32, requires_grad=False) if affine else None
-    self.bias = Tensor.zeros(sz, dtype=dtypes.float32, requires_grad=False) if affine else None
-
-    if track_running_stats: self.running_mean, self.running_var = Tensor.zeros(sz, dtype=dtypes.float32, requires_grad=False), Tensor.ones(sz, dtype=dtypes.float32, requires_grad=False)
-    self.num_batches_tracked = Tensor.zeros(1, dtype=dtypes.long, requires_grad=False)
-
-  def __call__(self, x:Tensor) -> Tensor:
-    batch_mean, batch_var = super().calc_stats(x.cast(dtypes.float32))
-    if self.track_running_stats and Tensor.training:
-      self.running_mean.assign((1-self.momentum) * self.running_mean + self.momentum * batch_mean.detach().cast(self.running_mean.dtype))
-      self.running_var.assign((1-self.momentum) * self.running_var + self.momentum * x.numel()/(x.numel()-x.shape[1]) * batch_var.detach().cast(self.running_var.dtype))
-      self.num_batches_tracked += 1
-    return x.cast(dtypes.float32).batchnorm(self.weight, self.bias, batch_mean, batch_var.add(self.eps).rsqrt()).cast(x.dtype)
-
-class Conv2dNormalRetinaNet(nn.Conv2d):
-  def __init__(self, in_channels:int, out_channels:int, kernel_size:int|tuple[int, ...],
-               stride:int=1, padding:int|tuple[int, ...]|str=0, dilation:int=1, groups:int=1,
-               bias:bool=True, prior_prob:float|None=None):
-    super().__init__(in_channels, out_channels, kernel_size, stride=stride, padding=padding, dilation=dilation, groups=groups, bias=bias)
-    self.weight = Tensor.normal(*self.weight.shape, std=0.01, dtype=dtypes.float32)
-    if bias:
-      if prior_prob:
-        prior_prob = Tensor(prior_prob, device=self.bias.device, dtype=dtypes.float32).expand(*self.bias.shape)
-        self.bias = -(((1 - prior_prob) / prior_prob).log())
-      else: self.bias = Tensor.zeros_like(self.bias, dtype=dtypes.float32)
-
-  def __call__(self, x:Tensor) -> Tensor:
-    return x.conv2d(self.weight.cast(dtypes.default_float), self.bias.cast(dtypes.default_float) if self.bias is not None else None,
-                    groups=self.groups, stride=self.stride, padding=self.padding)
-
-class Conv2dKaimingUniformRetinaNet(nn.Conv2d):
-  def __init__(self, in_channels:int, out_channels:int, kernel_size:int|tuple[int, ...],
-               stride:int=1, padding:int|tuple[int, ...]|str=0, dilation:int=1, groups:int=1,
-               bias:bool=True):
-    super().__init__(in_channels, out_channels, kernel_size, stride=stride, padding=padding, dilation=dilation, groups=groups, bias=bias)
-    self.weight = Tensor.kaiming_uniform(*self.weight.shape, a=1, dtype=dtypes.float32)
-    if bias: self.bias = Tensor.zeros_like(self.bias, dtype=dtypes.float32)
-
-  def __call__(self, x:Tensor) -> Tensor:
-    return x.conv2d(self.weight.cast(dtypes.default_float), self.bias.cast(dtypes.default_float) if self.bias is not None else None,
-                    groups=self.groups, stride=self.stride, padding=self.padding)
-
-class Conv2dRetinaNet(nn.Conv2d):
-  def __init__(self, in_channels:int, out_channels:int, kernel_size:int|tuple[int, ...],
-               stride:int=1, padding:int|tuple[int, ...]|str=0, dilation:int=1, groups:int=1,
-               bias:bool=True):
-    super().__init__(in_channels, out_channels, kernel_size, stride=stride, padding=padding, dilation=dilation, groups=groups, bias=bias)
-    scale = 1 / math.sqrt(in_channels * prod(self.kernel_size))
-    self.weight = Tensor.uniform(out_channels, in_channels//groups, *self.kernel_size, low=-scale, high=scale, dtype=dtypes.float32)
-    self.bias: Tensor|None = Tensor.uniform(out_channels, low=-scale, high=scale, dtype=dtypes.float32) if bias else None
-
-  def __call__(self, x:Tensor) -> Tensor:
-    return x.conv2d(self.weight.cast(dtypes.default_float), self.bias.cast(dtypes.default_float) if self.bias is not None else None,
-                    groups=self.groups, stride=self.stride, dilation=self.dilation, padding=self.padding)
@@ -1,29 +1,6 @@
 from examples.mlperf.metrics import dice_score
-from tinygrad import Tensor

 def dice_ce_loss(pred, tgt):
  ce = pred.permute(0, 2, 3, 4, 1).sparse_categorical_crossentropy(tgt.squeeze(1))
  dice = (1.0 - dice_score(pred, tgt, argmax=False, to_one_hot_x=False)).mean()
  return (dice + ce) / 2
-
-def sigmoid_focal_loss(pred:Tensor, tgt:Tensor, alpha:float=0.25, gamma:float=2.0, reduction:str="none") -> Tensor:
-  assert reduction in ["mean", "sum", "none"], f"unsupported reduction {reduction}"
-  p, ce_loss = pred.sigmoid(), pred.binary_crossentropy_logits(tgt, reduction="none")
-  p_t = p * tgt + (1 - p) * (1 - tgt)
-  loss = ce_loss * ((1 - p_t) ** gamma)
-
-  if alpha >= 0:
-    alpha_t = alpha * tgt + (1 - alpha) * (1 - tgt)
-    loss = loss * alpha_t
-
-  if reduction == "mean": loss = loss.mean()
-  elif reduction == "sum": loss = loss.sum()
-  return loss
-
-def l1_loss(pred:Tensor, tgt:Tensor, reduction:str="none") -> Tensor:
-  assert reduction in ["mean", "sum", "none"], f"unsupported reduction {reduction}"
-  loss = (pred - tgt).abs()
-
-  if reduction == "mean": loss = loss.mean()
-  elif reduction == "sum": loss = loss.sum()
-  return loss
@@ -81,43 +81,47 @@ def eval_unet3d():

 def eval_retinanet():
  # RetinaNet with ResNeXt50_32X4D
-  from examples.mlperf.dataloader import batch_load_retinanet
-  from extra.datasets.openimages import normalize, download_dataset, BASEDIR
  from extra.models.resnet import ResNeXt50_32X4D
  from extra.models.retinanet import RetinaNet
+  mdl = RetinaNet(ResNeXt50_32X4D())
+  mdl.load_from_pretrained()
+
+  input_mean = Tensor([0.485, 0.456, 0.406]).reshape(1, -1, 1, 1)
+  input_std = Tensor([0.229, 0.224, 0.225]).reshape(1, -1, 1, 1)
+  def input_fixup(x):
+    x = x.permute([0,3,1,2]) / 255.0
+    x -= input_mean
+    x /= input_std
+    return x
+
+  from extra.datasets.openimages import download_dataset, iterate, BASEDIR
  from pycocotools.coco import COCO
  from pycocotools.cocoeval import COCOeval
  from contextlib import redirect_stdout
-  tlog("imports")
-
-  mdl = RetinaNet(ResNeXt50_32X4D())
-  mdl.load_from_pretrained()
-  tlog("loaded models")
-
-  coco = COCO(download_dataset(base_dir:=getenv("BASEDIR", BASEDIR), 'validation'))
+  coco = COCO(download_dataset(base_dir:=getenv("BASE_DIR", BASEDIR), 'validation'))
  coco_eval = COCOeval(coco, iouType="bbox")
  coco_evalimgs, evaluated_imgs, ncats, narea = [], [], len(coco_eval.params.catIds), len(coco_eval.params.areaRng)
-  tlog("loaded dataset")

-  iterator = batch_load_retinanet(coco, True, Path(base_dir), getenv("BS", 8), shuffle=False)
-  def data_get():
-    x, img_ids, img_sizes, cookie = next(iterator)
-    return x.to(Device.DEFAULT).realize(), img_ids, img_sizes, cookie
-  n = 0
-  proc = data_get()
-  tlog("loaded initial data")
+  from tinygrad.engine.jit import TinyJit
+  mdlrun = TinyJit(lambda x: mdl(input_fixup(x)).realize())
+
+  n, bs = 0, 8
  st = time.perf_counter()
-  while proc is not None:
-    GlobalCounters.reset()
-    proc = (mdl(normalize(proc[0])), proc[1], proc[2], proc[3])
-    run = time.perf_counter()
-    # load the next data here
-    try: next_proc = data_get()
-    except StopIteration: next_proc = None
-    nd = time.perf_counter()
-    predictions, img_ids = mdl.postprocess_detections(proc[0].numpy(), orig_image_sizes=proc[2]), proc[1]
-    pd = time.perf_counter()
-    coco_results  = [{"image_id": img_ids[i], "category_id": label, "bbox": box.tolist(), "score": score}
+  for x, targets in iterate(coco, base_dir, bs):
+    dat = Tensor(x.astype(np.float32))
+    mt = time.perf_counter()
+    if dat.shape[0] == bs:
+      outs = mdlrun(dat).numpy()
+    else:
+      mdlrun._jit_cache = []
+      outs =  mdl(input_fixup(dat)).numpy()
+    et = time.perf_counter()
+    predictions = mdl.postprocess_detections(outs, input_size=dat.shape[1:3], orig_image_sizes=[t["image_size"] for t in targets])
+    ext = time.perf_counter()
+    n += len(targets)
+    print(f"[{n}/{len(coco.imgs)}] == {(mt-st)*1000:.2f} ms loading data, {(et-mt)*1000:.2f} ms to run model, {(ext-et)*1000:.2f} ms for postprocessing")
+    img_ids = [t["image_id"] for t in targets]
+    coco_results  = [{"image_id": targets[i]["image_id"], "category_id": label, "bbox": box.tolist(), "score": score}
      for i, prediction in enumerate(predictions) for box, score, label in zip(*prediction.values())]
    with redirect_stdout(None):
      coco_eval.cocoDt = coco.loadRes(coco_results)
@@ -125,18 +129,13 @@ def eval_retinanet():
      coco_eval.evaluate()
    evaluated_imgs.extend(img_ids)
    coco_evalimgs.append(np.array(coco_eval.evalImgs).reshape(ncats, narea, len(img_ids)))
-    n += len(proc[0])
-    et = time.perf_counter()
-    tlog(f"****** {(run-st)*1000:7.2f} ms to enqueue, {(et-run)*1000:7.2f} ms to realize ({(nd-run)*1000:7.2f} ms fetching, {(pd-run)*1000:4.2f} ms postprocess_detections). {(len(proc))/(et-st):8.2f} examples/sec. {GlobalCounters.global_ops*1e-12/(et-st):5.2f} TFLOPS")
-    st = et
-    proc, next_proc = next_proc, None
+    st = time.perf_counter()

  coco_eval.params.imgIds = evaluated_imgs
  coco_eval._paramsEval.imgIds = evaluated_imgs
  coco_eval.evalImgs = list(np.concatenate(coco_evalimgs, -1).flatten())
  coco_eval.accumulate()
  coco_eval.summarize()
-  tlog("done")

 def eval_rnnt():
  # RNN-T
@@ -1,11 +1,11 @@
-import os, time, math, functools, random
+import os, time, math, functools
 from pathlib import Path
 import multiprocessing

 from tinygrad import Device, GlobalCounters, Tensor, TinyJit, dtypes
-from tinygrad.helpers import getenv, BEAM, WINO, round_up, diskcache_clear, FUSE_CONV_BW, Profiling
+from tinygrad.helpers import getenv, BEAM, WINO, round_up, diskcache_clear, FUSE_CONV_BW
 from tinygrad.nn.state import get_parameters, get_state_dict, safe_load, safe_save
-from tinygrad.nn.optim import LAMB, LARS, SGD, OptimizerGroup, Adam
+from tinygrad.nn.optim import LAMB, LARS, SGD, OptimizerGroup

 from extra.lr_scheduler import LRSchedulerGroup
 from examples.mlperf.helpers import get_training_state, load_training_state
@@ -343,348 +343,8 @@ def train_resnet():
        safe_save(get_training_state(model, optimizer_group, scheduler_group), fn)

 def train_retinanet():
-  from contextlib import redirect_stdout
-  from examples.mlperf.dataloader import batch_load_retinanet
-  from examples.mlperf.initializers import FrozenBatchNorm2dRetinaNet, Conv2dNormalRetinaNet, Conv2dKaimingUniformRetinaNet, Linear, Conv2dRetinaNet
-  from extra.datasets.openimages import MLPERF_CLASSES, BASEDIR, download_dataset, normalize, get_dataset_count
-  from extra.models import resnet
-  from pycocotools.coco import COCO
-  from pycocotools.cocoeval import COCOeval
-  from tinygrad.helpers import colored
-  from typing import Iterator
-  import extra.models.retinanet as retinanet
-
-  import numpy as np
-
-  config, target_metric = {}, 0.34
-
-  config["SEED"] = SEED = getenv("SEED", random.SystemRandom().randint(0, 2**32 - 1))
-  Tensor.manual_seed(SEED)
-
-  NUM_CLASSES = len(MLPERF_CLASSES)
-  BASEDIR = getenv("BASEDIR", BASEDIR)
-  BENCHMARK = getenv("BENCHMARK")
-  INITMLPERF = getenv("INITMLPERF")
-  RUNMLPERF = getenv("RUNMLPERF")
-
-  if getenv("LOGMLPERF"):
-    from mlperf_logging import mllog
-    import mlperf_logging.mllog.constants as mllog_constants
-
-    mllog.config(filename=f"result_retinanet_{SEED}.log")
-    mllog.config(root_dir=Path(__file__).parents[3].as_posix())
-    MLLOGGER = mllog.get_mllogger()
-    MLLOGGER.logger.propagate = False
-
-    if INITMLPERF:
-      assert BENCHMARK, "BENCHMARK must be set for INITMLPERF"
-      MLLOGGER.event(key=mllog_constants.SUBMISSION_ORG, value="tinycorp")
-      MLLOGGER.event(key=mllog_constants.SUBMISSION_PLATFORM, value=getenv("SUBMISSION_PLATFORM", "tinybox"))
-      MLLOGGER.event(key=mllog_constants.SUBMISSION_DIVISION, value=mllog_constants.CLOSED)
-      MLLOGGER.event(key=mllog_constants.SUBMISSION_STATUS, value=mllog_constants.ONPREM)
-
-      MLLOGGER.event(key=mllog_constants.SUBMISSION_BENCHMARK, value=mllog_constants.RETINANET)
-
-      diskcache_clear()
-      MLLOGGER.event(key=mllog_constants.CACHE_CLEAR, value=True)
-      MLLOGGER.start(key=mllog_constants.INIT_START)
-
-    if RUNMLPERF:
-      MLLOGGER.start(key=mllog_constants.RUN_START)
-      MLLOGGER.event(key=mllog_constants.SEED, value=SEED)
-  else:
-    MLLOGGER = None
-
-  config["gpus"] = GPUS = [f"{Device.DEFAULT}:{i}" for i in range(getenv("GPUS", 6))]
-
-  for x in GPUS: Device[x]
-  print(f"training on {GPUS}")
-
-  def _freeze_backbone_layers(backbone:resnet.ResNet, trainable_layers:int):
-    layers_to_train = ["layer4", "layer3", "layer2", "layer1", "conv1"][:trainable_layers]
-    for k, v in get_state_dict(backbone).items():
-      if all([not k.startswith(layer) for layer in layers_to_train]):
-        v.requires_grad = False
-
-  def _data_get(it:Iterator[tuple[Tensor, ...]], val:bool=False):
-    if val:
-      x, img_ids, img_sizes, cookie = next(it)
-      return x.shard(GPUS, axis=0), img_ids, img_sizes, cookie
-
-    x, y_boxes, y_labels, matches, anchors, cookie = next(it)
-    return x.shard(GPUS, axis=0), y_boxes.shard(GPUS, axis=0), y_labels.shard(GPUS, axis=0), matches.shard(GPUS, axis=0), anchors.shard(GPUS, axis=0), cookie
-
-  def _fake_data_get(bs:int, val:bool=False):
-    x = Tensor.empty(bs, 800, 800, 3, dtype=dtypes.uint8)
-    if val:
-      img_ids, img_sizes = [0] * bs, [(800, 800)] * bs
-      return x.shard(GPUS, axis=0), img_ids, img_sizes, None
-
-    y_boxes = Tensor.empty(bs, 120087, 4, dtype=dtypes.float32)
-    y_labels = Tensor.empty(bs, 120087, dtype=dtypes.int64)
-    matches = Tensor.empty(bs, 120087, dtype=dtypes.int64)
-    anchors = Tensor.empty(bs, 120087, 4, dtype=dtypes.float64)
-    return x.shard(GPUS, axis=0), y_boxes.shard(GPUS, axis=0), y_labels.shard(GPUS, axis=0), matches.shard(GPUS, axis=0), anchors.shard(GPUS, axis=0), None
-
-  @TinyJit
-  def _train_step(model, optim, loss_scaler, x, **kwargs):
-    optim.zero_grad()
-
-    losses = model(normalize(x, GPUS), **kwargs)
-    loss = sum([l for l in losses.values()])
-
-    (loss * loss_scaler).backward()
-    for t in optim.params: t.grad = t.grad / loss_scaler
-
-    optim.step()
-
-    return loss.realize(), losses
-
-  @TinyJit
-  def _eval_step(model, x, **kwargs):
-    out = model(normalize(x, GPUS), **kwargs)
-    # reassemble on GPUS[0] before sending back to CPU for speed
-    return out.to(GPUS[0]).realize()
-
-  # ** hyperparameters **
-  config["BS"] = BS = getenv("BS", 16 * len(GPUS) if dtypes.default_float == dtypes.float16 else 12 * len(GPUS))
-  config["EVAL_BS"] = EVAL_BS = getenv("EVAL_BS", BS)
-  config["EPOCHS"] = EPOCHS = getenv("EPOCHS", 4)
-  config["TRAIN_BEAM"] = TRAIN_BEAM = getenv("TRAIN_BEAM", BEAM.value)
-  config["EVAL_BEAM"] = EVAL_BEAM = getenv("EVAL_BEAM", BEAM.value)
-  config["LR"] = lr = getenv("LR", 9.5e-5 * (BS / 96))
-  config["LOSS_SCALER"] = loss_scaler = getenv("LOSS_SCALER", 2**11 if dtypes.default_float == dtypes.float16 else 1.0)
-  config["DEFAULT_FLOAT"] = dtypes.default_float.name
-  config["EVAL_FREQ"] = eval_freq = getenv("EVAL_FREQ", 1)
-
-  # ** initialize wandb **
-  if (WANDB:=getenv("WANDB")):
-    import wandb
-    wandb.init(config=config, project="MLPerf-RetinaNet")
-
-  # ** model initializers **
-  resnet.BatchNorm = FrozenBatchNorm2dRetinaNet
-  resnet.Linear = Linear
-  resnet.Conv2d = Conv2dRetinaNet
-
-  retinanet.ConvHead = Conv2dNormalRetinaNet
-  retinanet.ConvClassificationHeadLogits = functools.partial(Conv2dNormalRetinaNet, prior_prob=0.01)
-  retinanet.ConvFPN = Conv2dKaimingUniformRetinaNet
-
-  # ** model setup **
-  backbone = resnet.ResNeXt50_32X4D(num_classes=None)
-  if RUNMLPERF:
-    backbone.load_from_pretrained()
-  _freeze_backbone_layers(backbone, 3)
-
-  model = retinanet.RetinaNet(backbone, num_classes=NUM_CLASSES)
-  params = get_parameters(model)
-
-  if not RUNMLPERF:
-    # for init, zero out all weights
-    for p in params:
-      p = p.assign(Tensor.zeros_like(p).contiguous()).realize()
-
-  if len(GPUS) > 1:
-    for p in params: p.to_(GPUS)
-
-  step_times, start_epoch = [], 0
-
-  # ** optimizer **
-  optim = Adam(params, lr=lr)
-
-  # ** dataset **
-  config["STEPS_IN_TRAIN_EPOCH"] = steps_in_train_epoch = round_up(get_dataset_count((base_dir_path:=Path(BASEDIR)), False), BS) // BS
-  config["STEPS_IN_VAL_EPOCH"] = steps_in_val_epoch = (round_up(get_dataset_count(base_dir_path, True), EVAL_BS) // EVAL_BS)
-
-  # log mlperf hparams
-  if MLLOGGER:
-    if RUNMLPERF:
-      MLLOGGER.event(key=mllog_constants.GLOBAL_BATCH_SIZE, value=config["BS"])
-      MLLOGGER.event(key=mllog_constants.TRAIN_SAMPLES, value=config["STEPS_IN_TRAIN_EPOCH"])
-      MLLOGGER.event(key=mllog_constants.EVAL_SAMPLES, value=config["STEPS_IN_VAL_EPOCH"])
-      MLLOGGER.event(key=mllog_constants.EPOCH_COUNT, value=config["EPOCHS"])
-      MLLOGGER.event(key=mllog_constants.FIRST_EPOCH_NUM, value=start_epoch)
-
-      MLLOGGER.event(key=mllog_constants.OPT_NAME, value=mllog_constants.ADAM)
-      MLLOGGER.event(key=mllog_constants.OPT_BASE_LR, value=config["LR"])
-      MLLOGGER.event(key=mllog_constants.OPT_WEIGHT_DECAY, value=0)
-      MLLOGGER.event(key=mllog_constants.OPT_LR_WARMUP_EPOCHS, value=0)
-      MLLOGGER.event(key=mllog_constants.OPT_LR_WARMUP_FACTOR, value=0)
-      MLLOGGER.event(key=mllog_constants.GRADIENT_ACCUMULATION_STEPS, value=1)
-
-  if RUNMLPERF:
-    train_dataset = COCO(download_dataset(BASEDIR, "train"))
-    val_dataset = COCO(download_dataset(BASEDIR, "validation"))
-    coco_val = COCOeval(cocoGt=val_dataset, iouType="bbox")
-
-  print(f"training with batch size {BS} for {EPOCHS} epochs")
-
-  for e in range(start_epoch, EPOCHS):
-    # ** training loop **
-    if MLLOGGER and RUNMLPERF:
-      MLLOGGER.start(key=mllog_constants.EPOCH_START, value=e + 1, metadata={"epoch_num": e + 1})
-
-    BEAM.value = TRAIN_BEAM
-
-    if not RUNMLPERF:
-      i, proc = 0, _fake_data_get(BS)
-    else:
-      train_dataloader = batch_load_retinanet(train_dataset, False, base_dir_path, batch_size=BS, seed=SEED)
-      it = iter(tqdm(train_dataloader, total=steps_in_train_epoch, desc=f"epoch {e + 1}", disable=BENCHMARK))
-      i, proc = 0, _data_get(it)
-
-    prev_cookies = []
-    st = time.perf_counter()
-
-    while proc is not None:
-      GlobalCounters.reset()
-
-      x, y_bboxes, y_labels, matches, anchors, proc = proc
-      loss, losses = _train_step(model, optim, loss_scaler, x, labels=y_labels, matches=matches, anchors=anchors, bboxes=y_bboxes)
-
-      pt = time.perf_counter()
-
-      if len(prev_cookies) == getenv("STORE_COOKIES", 1): prev_cookies = []  # free previous cookies after gpu work has been enqueued
-      try:
-        if not RUNMLPERF:
-          next_proc = _fake_data_get(BS)
-        else:
-          next_proc = _data_get(it)
-      except StopIteration:
-        next_proc = None
-
-      dt = time.perf_counter()
-
-      device_str = loss.device if isinstance(loss.device, str) else f"{loss.device[0]} * {len(loss.device)}"
-      loss = loss.item()
-
-      cl = time.perf_counter()
-      if BENCHMARK: step_times.append(cl - st)
-
-      if not math.isfinite(loss):
-        print("loss is nan")
-        return
-
-      tqdm.write(
-        f"{i:5} {((cl - st)) * 1000.0:7.2f} ms run, {(pt - st) * 1000.0:7.2f} ms python, {(dt - pt) * 1000.0:6.2f} ms fetch data, "
-        f"{(cl - dt) * 1000.0:7.2f} ms {device_str}, {loss:5.2f} loss, {losses['classification_loss'].item():5.4f} classification loss, {losses['regression_loss'].item():5.4f} regression loss, "
-        f"{optim.lr.numpy()[0]:.6f} LR, {GlobalCounters.mem_used / 1e9:.2f} GB used, {GlobalCounters.global_ops * 1e-9 / (cl - st):9.2f} GFLOPS"
-      )
-
-      if WANDB:
-        wandb.log({"lr": optim.lr.numpy(), "train/loss": loss, "train/classification_loss": losses["classification_loss"].item(), "train/regression_loss": losses["regression_loss"].item(),
-                  "train/step_time": cl - st, "train/python_time": pt - st, "train/data_time": dt - pt, "train/cl_time": cl - dt,
-                  "train/GFLOPS": GlobalCounters.global_ops * 1e-9 / (cl - st), "epoch": e + (i + 1) / steps_in_train_epoch})
-
-      st = cl
-      prev_cookies.append(proc)
-      proc, next_proc = next_proc, None  # return old cookie
-      i += 1
-
-      if i == BENCHMARK:
-        assert not math.isnan(loss)
-        median_step_time = sorted(step_times)[(BENCHMARK + 1) // 2]  # in seconds
-        estimated_total_minutes = int(median_step_time * steps_in_train_epoch * EPOCHS / 60)
-        print(f"Estimated training time: {estimated_total_minutes // 60}h{estimated_total_minutes % 60}m")
-        print(f"epoch global_ops: {steps_in_train_epoch * GlobalCounters.global_ops:_}, "
-              f"epoch global_mem: {steps_in_train_epoch * GlobalCounters.global_mem:_}")
-        # if we are doing beam search, run the first eval too
-        if (TRAIN_BEAM or EVAL_BEAM) and e == start_epoch: break
-        return
-
-    if MLLOGGER and RUNMLPERF:
-      MLLOGGER.event(key=mllog_constants.EPOCH_STOP, value=e + 1, metadata={"epoch_num": e + 1})
-
-    # ** eval loop **
-    if (e + 1) % eval_freq == 0:
-      if MLLOGGER and RUNMLPERF:
-        MLLOGGER.start(key=mllog_constants.EVAL_START, value=e + 1, metadata={"epoch_num": e + 1})
-
-      BEAM.value = EVAL_BEAM
-
-      if getenv("RESET_STEP", 1): _train_step.reset()
-
-      with Tensor.train(mode=False), Tensor.test():
-        if not RUNMLPERF:
-          i, proc = 0, _fake_data_get(EVAL_BS, val=(val:=True))
-        else:
-          val_dataloader = batch_load_retinanet(val_dataset, (val:=True), Path(BASEDIR), batch_size=EVAL_BS, shuffle=False, seed=SEED)
-          it = iter(tqdm(val_dataloader, total=steps_in_val_epoch))
-          i, proc = 0, _data_get(it, val=val)
-          val_img_ids, val_imgs, ncats, narea = [], [], len(coco_val.params.catIds), len(coco_val.params.areaRng)
-
-        eval_times, prev_cookies = [], []
-
-        while proc is not None:
-          GlobalCounters.reset()
-          st = time.time()
-
-          out, img_ids, img_sizes, proc = _eval_step(model, (x:=proc[0])).numpy(), proc[1], proc[2], proc[3]
-
-          if RUNMLPERF:
-            out = model.postprocess_detections(out, input_size=x.shape[1:3], orig_image_sizes=img_sizes)
-            coco_results  = [{"image_id": img_ids[i], "category_id": label, "bbox": box.tolist(), "score": score}
-              for i, prediction in enumerate(out) for box, score, label in zip(*prediction.values())]
-
-            with redirect_stdout(None):
-              coco_val.cocoDt = val_dataset.loadRes(coco_results)
-              coco_val.params.imgIds = img_ids
-              coco_val.evaluate()
-
-            val_img_ids.extend(img_ids)
-            val_imgs.append(np.array(coco_val.evalImgs).reshape(ncats, narea, len(img_ids)))
-
-          if len(prev_cookies) == getenv("STORE_COOKIES", 1): prev_cookies = []  # free previous cookies after gpu work has been enqueued
-          try:
-            if not RUNMLPERF:
-              next_proc = _fake_data_get(EVAL_BS, val=val)
-            else:
-              next_proc = _data_get(it, val=val)
-          except StopIteration:
-            next_proc = None
-
-          prev_cookies.append(proc)
-          proc, next_proc = next_proc, None
-          i += 1
-
-          et = time.time()
-          eval_times.append(et - st)
-
-          if i == BENCHMARK:
-            # assume INITMLPERF has BENCHMARK set
-            if MLLOGGER and INITMLPERF:
-              MLLOGGER.event(key=mllog_constants.INIT_STOP)
-            return
-
-        if getenv("RESET_STEP", 1): _eval_step.reset()
-        total_fw_time = sum(eval_times) / len(eval_times)
-
-        if RUNMLPERF:
-          coco_val.params.imgIds = val_img_ids
-          coco_val._paramsEval.imgIds = val_img_ids
-          coco_val.evalImgs = list(np.concatenate(val_imgs, -1).flatten())
-          coco_val.accumulate()
-          coco_val.summarize()
-
-          val_metric = coco_val.stats[0]
-
-          tqdm.write(f"eval time: {total_fw_time:.2f}, eval metric: {val_metric:.4f}")
-
-          if WANDB:
-            wandb.log({"eval/forward_time": total_fw_time, "eval/metric": val_metric, "epoch": e + 1})
-
-          if MLLOGGER:
-            MLLOGGER.event(key=mllog_constants.EVAL_ACCURACY, value=val_metric, metadata={"epoch_num": e + 1}, clear_line=True)
-            MLLOGGER.end(key=mllog_constants.EVAL_STOP, value=e + 1, metadata={"epoch_num": e + 1})
-
-          if val_metric >= target_metric:
-            print(colored(f"target metric reached: {val_metric:.2f}/{target_metric:.2f}", color="green"))
-
-            if MLLOGGER:
-              MLLOGGER.end(key=mllog_constants.RUN_STOP, metadata={"status": mllog_constants.SUCCESS})
-
-            break
+  # TODO: Retinanet
+  pass

 def train_unet3d():
  """
@@ -923,13 +583,12 @@ def train_step_bert(model, optimizer, scheduler, loss_scaler:float, input_ids:Te
  loss = model.loss(lm_logits, seq_relationship_logits, masked_lm_ids, masked_lm_weights, next_sentence_labels)
  (loss * loss_scaler).backward()

-  global_norm = Tensor([0.0], dtype=dtypes.float32, device=optimizer[0].device)
+  global_norm = Tensor([0.0], dtype=dtypes.float32, device=optimizer[0].device).realize()
  for p in optimizer.params:
    p.grad = p.grad / loss_scaler
    global_norm += p.grad.float().square().sum()
-  global_norm = global_norm.sqrt().contiguous()
-  for p in optimizer.params:
-    p.grad = (global_norm > 1.0).where((p.grad/global_norm).cast(p.grad.dtype), p.grad)
+  global_norm = global_norm.sqrt()
+  for p in optimizer.params: p.grad = (p.grad / Tensor.where(global_norm > 1.0, global_norm, 1.0)).cast(p.grad.dtype)

  optimizer.step()
  scheduler.step()
@@ -1000,10 +659,8 @@ def train_bert():
  BS                 = config["GLOBAL_BATCH_SIZE"]      = getenv("BS", 11 * len(GPUS) if dtypes.default_float in (dtypes.float16, dtypes.bfloat16) else 8 * len(GPUS))
  EVAL_BS            = config["EVAL_BS"]                = getenv("EVAL_BS", 1 * len(GPUS))
  max_lr             = config["OPT_BASE_LEARNING_RATE"] = getenv("OPT_BASE_LEARNING_RATE", 0.000175 * math.sqrt(BS/96))
-  opt_lamb_beta_1    = config["OPT_LAMB_BETA_1"]        = getenv("OPT_LAMB_BETA_1", 0.9)
-  opt_lamb_beta_2    = config["OPT_LAMB_BETA_2"]        = getenv("OPT_LAMB_BETA_2", 0.999)

-  train_steps        = config["TRAIN_STEPS"]            = getenv("TRAIN_STEPS", 3600000 // BS)
+  train_steps        = config["TRAIN_STEPS"]            = getenv("TRAIN_STEPS", 3300000 // BS)
  warmup_steps       = config["NUM_WARMUP_STEPS"]       = getenv("NUM_WARMUP_STEPS", 1)
  max_eval_steps     = config["MAX_EVAL_STEPS"]         = getenv("MAX_EVAL_STEPS", (10000 + EVAL_BS - 1) // EVAL_BS) # EVAL_BS * MAX_EVAL_STEPS >= 10000
  eval_step_freq     = config["EVAL_STEP_FREQ"]         = getenv("EVAL_STEP_FREQ", int((math.floor(0.05 * (230.23 * BS + 3000000) / 25000) * 25000) / BS)) # Round down
@@ -1056,8 +713,8 @@ def train_bert():
  # ** Optimizer **
  parameters_no_wd = [v for k, v in get_state_dict(model).items() if "bias" in k or "LayerNorm" in k]
  parameters = [x for x in parameters if x not in set(parameters_no_wd)]
-  optimizer_wd = LAMB(parameters, lr=max_lr, b1=opt_lamb_beta_1, b2=opt_lamb_beta_2, eps=epsilon, weight_decay=decay, adam=False)
-  optimizer_no_wd = LAMB(parameters_no_wd, lr=max_lr, b1=opt_lamb_beta_1, b2=opt_lamb_beta_2, eps=epsilon, weight_decay=0.0, adam=False)
+  optimizer_wd = LAMB(parameters, lr=max_lr, eps=epsilon, weight_decay=decay, adam=False)
+  optimizer_no_wd = LAMB(parameters_no_wd, lr=max_lr, eps=epsilon, weight_decay=0.0, adam=False)
  optimizer_group = OptimizerGroup(optimizer_wd, optimizer_no_wd)

  # ** LR scheduler **
@@ -1076,8 +733,8 @@ def train_bert():
      MLLOGGER.event(key=mllog_constants.OPT_NAME, value="LAMB")
      MLLOGGER.event(key=mllog_constants.OPT_BASE_LR, value=config["OPT_BASE_LEARNING_RATE"])
      MLLOGGER.event(key=mllog_constants.OPT_LAMB_WEIGHT_DECAY, value=config["DECAY"])
-      MLLOGGER.event(key=mllog_constants.OPT_LAMB_BETA_1, value=config["OPT_LAMB_BETA_1"])
-      MLLOGGER.event(key=mllog_constants.OPT_LAMB_BETA_2, value=config["OPT_LAMB_BETA_2"])
+      MLLOGGER.event(key=mllog_constants.OPT_LAMB_BETA_1, value=optimizer_wd.b1)
+      MLLOGGER.event(key=mllog_constants.OPT_LAMB_BETA_2, value=optimizer_wd.b2)
      MLLOGGER.event(key=mllog_constants.OPT_LAMB_LR_DECAY_POLY_POWER, value=config["POLY_POWER"])
      MLLOGGER.event(key=mllog_constants.OPT_LAMB_EPSILON, value=config["EPSILON"])

@@ -1140,7 +797,6 @@ def train_bert():

      device_str = parameters[0].device if isinstance(parameters[0].device, str) else f"{parameters[0].device[0]} * {len(parameters[0].device)}"
      loss = loss.item()
-      assert not math.isnan(loss)
      lr = lr.item()

      cl = time.perf_counter()
@@ -1169,8 +825,8 @@ def train_bert():
    if i % eval_step_freq == 0 or (BENCHMARK and i == BENCHMARK) or i == train_steps:
      if MLLOGGER and RUNMLPERF:
        MLLOGGER.start(key=mllog_constants.EVAL_START, value=None, metadata={"epoch_num": i*BS, "step_num": i})
-      if getenv("RESET_STEP"): train_step_bert.reset()
-      elif getenv("FREE_INTERMEDIATE", 1) and train_step_bert.captured is not None: train_step_bert.captured.free_intermediates()
+      if getenv("RESET_STEP", 0): train_step_bert.reset()
+      else: train_step_bert.captured.free_intermediates()
      eval_lm_losses = []
      eval_clsf_losses = []
      eval_lm_accs = []
@@ -1197,15 +853,14 @@ def train_bert():
        et = time.time()
        eval_times.append(et - st)

-        if BENCHMARK and (j+1) == min(BENCHMARK, max_eval_steps):
+        if BENCHMARK and j == BENCHMARK:
          # assume INITMLPERF has BENCHMARK set
          if MLLOGGER and INITMLPERF:
            MLLOGGER.event(key=mllog_constants.INIT_STOP, value=None)
          return

-      if getenv("RESET_STEP"): eval_step_bert.reset()
-      elif getenv("FREE_INTERMEDIATE", 1) and eval_step_bert.captured is not None: eval_step_bert.captured.free_intermediates()
-
+      if getenv("RESET_STEP", 0): eval_step_bert.reset()
+      else: eval_step_bert.captured.free_intermediates()
      del eval_data
      avg_lm_loss = sum(eval_lm_losses) / len(eval_lm_losses)
      avg_clsf_loss = sum(eval_clsf_losses) / len(eval_clsf_losses)
@@ -1218,7 +873,7 @@ def train_bert():

      if WANDB:
        wandb.log({"eval/lm_loss": avg_lm_loss, "eval/clsf_loss": avg_clsf_loss, "eval/lm_accuracy": avg_lm_acc, \
-                    "eval/clsf_accuracy": avg_clsf_acc, "eval/forward_time": avg_fw_time, "epoch": (i+1)*BS})
+                    "eval/clsf_accuracy": avg_clsf_acc, "eval/forward_time": avg_fw_time})

      if MLLOGGER and RUNMLPERF:
        MLLOGGER.end(key=mllog_constants.EVAL_STOP, value=i*BS, metadata={"epoch_count": i*BS, "step_num": i, "samples_count": config["EVAL_BS"] * config["MAX_EVAL_STEPS"]})
@@ -1245,9 +900,6 @@ def train_bert():
        # stop once hitting the target
        break

-    # should not happen, BENCHMARK not properly terminated
-    if BENCHMARK: assert i < BENCHMARK, i
-
    if getenv("CKPT") and i % save_ckpt_freq == 0:
      if MLLOGGER and RUNMLPERF:
        if previous_step:
@@ -1282,4 +934,4 @@ if __name__ == "__main__":
      nm = f"train_{m}"
      if nm in globals():
        print(f"training {m}")
-        with Profiling(enabled=getenv("PYPROFILE")): globals()[nm]()
+        globals()[nm]()
@@ -1,69 +0,0 @@
-# 1. Problem
-
-This problem uses BERT for NLP.
-
-## Requirements
-
-Install tinygrad and mlperf-logging (uncomment mlperf from setup.py) from branch mlperf_training_v5.0.
-```
-git clone https://github.com/tinygrad/tinygrad.git
-python3 -m pip install -e ".[mlperf]"
-```
-Also install gdown (for dataset), numpy, tqdm and tensorflow.
-```
-pip install gdown numpy tqdm tensorflow
-```
-
-### tinybox_green
-Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
-This is the default on production tinybox green.
-
-# 2. Directions
-
-## Steps to download and verify data
-
-### 1. Download raw data
-
-```
-BASEDIR="/raid/datasets/wiki" WIKI_TRAIN=1 VERIFY_CHECKSUM=1 python3 extra/datasets/wikipedia_download.py
-```
-
-### 2. Preprocess train and validation data
-
-Note: The number of threads used for preprocessing is limited by available memory. With 128GB of RAM, a maximum of 16 threads is recommended. 
-
-#### Training:
-```
-BASEDIR="/raid/datasets/wiki" NUM_WORKERS=16 python3 extra/datasets/wikipedia.py pre-train all
-```
-
-Generating a specific topic (Between 0 and 499)
-```
-BASEDIR="/raid/datasets/wiki" python3 extra/datasets/wikipedia.py pre-train 42
-```
-
-#### Validation:
-```
-BASEDIR="/raid/datasets/wiki" python3 extra/datasets/wikipedia.py pre-eval
-```
-## Running
-
-### tinybox_green
-
-#### Steps to run benchmark
-```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_green/run_and_time.sh
-```
-
-### tinybox_red
-
-#### Steps to run benchmark
-```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_red/run_and_time.sh
-```
-### tinybox_8xMI300X
-
-#### Steps to run benchmark
-```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_8xMI300X/run_and_time.sh
-```
@@ -1,14 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." AMD=1
-export MODEL="bert"
-export DEFAULT_FLOAT="HALF" GPUS=8 BS=1024 EVAL_BS=1024
-export OPT_BASE_LEARNING_RATE=0.0011 OPT_LAMB_BETA_1=0.60466 OPT_LAMB_BETA_2=0.85437 DECAY=0.1
-
-export BEAM=3 BEAM_UOPS_MAX=6000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
-export IGNORE_JIT_FIRST_BEAM=1 FREE_INTERMEDIATE=0
-export BASEDIR="/raid/datasets/wiki"
-
-export BENCHMARK=10 BERT_LAYERS=2 DEBUG=2
-
-python3 examples/mlperf/model_train.py
@@ -1,17 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." AMD=1
-export MODEL="bert"
-export DEFAULT_FLOAT="HALF" GPUS=8 BS=1024 EVAL_BS=1024
-
-# similar to https://github.com/mlcommons/training_results_v3.1/blob/d06288b2bd675a9d88e0e6181f5bb5626b71ec19/Quanta_Cloud_Technology/results/D54U-3U/bert/result_1.txt#L54
-export OPT_BASE_LEARNING_RATE=0.0011 OPT_LAMB_BETA_1=0.60466 OPT_LAMB_BETA_2=0.85437 DECAY=0.1
-export TRAIN_STEPS=3900
-
-export BEAM=3 BEAM_UOPS_MAX=6000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
-export IGNORE_JIT_FIRST_BEAM=1 FREE_INTERMEDIATE=0
-export BASEDIR="/raid/datasets/wiki"
-
-export WANDB=1 PARALLEL=0
-
-RUNMLPERF=1 python3 examples/mlperf/model_train.py
@@ -1,27 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." AMD=1
-export MODEL="bert"
-export SUBMISSION_PLATFORM="tinybox_8xMI300X"
-export DEFAULT_FLOAT="HALF" GPUS=8 BS=1024 EVAL_BS=1024
-
-# similar to https://github.com/mlcommons/training_results_v3.1/blob/d06288b2bd675a9d88e0e6181f5bb5626b71ec19/Quanta_Cloud_Technology/results/D54U-3U/bert/result_1.txt#L54
-export OPT_BASE_LEARNING_RATE=0.0011 OPT_LAMB_BETA_1=0.60466 OPT_LAMB_BETA_2=0.85437 DECAY=0.1
-export TRAIN_STEPS=3900
-
-export BEAM=3 BEAM_UOPS_MAX=6000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
-export IGNORE_JIT_FIRST_BEAM=1 FREE_INTERMEDIATE=0
-export BASEDIR="/raid/datasets/wiki"
-
-# pip install -e ".[mlperf]"
-export LOGMLPERF=1
-
-export SEED=$RANDOM
-DATETIME=$(date "+%m%d%H%M")
-LOGFILE="bert_8xMI300x_${DATETIME}_${SEED}.log"
-
-# init  # TODO: without DEBUG=2 it hangs
-BENCHMARK=10 INITMLPERF=1 BERT_LAYERS=2 DEBUG=2 python3 examples/mlperf/model_train.py | tee $LOGFILE
-
-# run
-PARALLEL=0 RUNMLPERF=1 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -4,20 +4,24 @@ This problem uses BERT for NLP.

 ## Requirements

-Install tinygrad and mlperf-logging (uncomment mlperf from setup.py) from branch mlperf_training_v5.0.
+Install tinygrad and mlperf-logging from master.
 ```
 git clone https://github.com/tinygrad/tinygrad.git
 python3 -m pip install -e ".[mlperf]"
 ```
-Also install gdown (for dataset), numpy, tqdm and tensorflow.
+Also install tqdm and tensorflow.
 ```
-pip install gdown numpy tqdm tensorflow
+pip install tqdm tensorflow
 ```

 ### tinybox_green
 Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
 This is the default on production tinybox green.

+### tinybox_red
+Disable cwsr + increase mes timeout.
+Install the custom amdgpu driver per [README](https://github.com/nimlgen/amdgpu_ubuntu_22_04/blob/v6.1.3/readme.md)
+
 # 2. Directions

 ## Steps to download and verify data
@@ -52,18 +56,18 @@ BASEDIR="/raid/datasets/wiki" python3 extra/datasets/wikipedia.py pre-eval

 #### Steps to run benchmark
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_green/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_green/run_and_time.sh
 ```

 ### tinybox_red

-#### Steps to run benchmark
+#### One time setup
+
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_red/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_red/setup.sh
 ```
-### tinybox_8xMI300X

 #### Steps to run benchmark
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_8xMI300X/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_red/run_and_time.sh
 ```
@@ -1,16 +1,14 @@
 #!/bin/bash

-export PYTHONPATH="." NV=1
+export PYTHONPATH="."
 export MODEL="bert"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=8 BEAM_UOPS_MAX=10000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=4 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
-export BEAM_LOG_SURPASS_MAX=1
 export BASEDIR="/raid/datasets/wiki"

-export BENCHMARK=10 BERT_LAYERS=2 DEBUG=2
+export RESET_STEP=1
+export BENCHMARK=10 DEBUG=2

 python3 examples/mlperf/model_train.py
@@ -1,12 +1,10 @@
 #!/bin/bash

-export PYTHONPATH="." NV=1
+export PYTHONPATH="."
 export MODEL="bert"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=8 BEAM_UOPS_MAX=10000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=4 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
 export BASEDIR="/raid/datasets/wiki"

@@ -1,13 +1,11 @@
 #!/bin/bash

-export PYTHONPATH="." NV=1
+export PYTHONPATH="."
 export MODEL="bert"
 export SUBMISSION_PLATFORM="tinybox_green"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=8 BEAM_UOPS_MAX=10000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=4 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
 export BASEDIR="/raid/datasets/wiki"

@@ -19,7 +17,7 @@ DATETIME=$(date "+%m%d%H%M")
 LOGFILE="bert_green_${DATETIME}_${SEED}.log"

 # init
-BENCHMARK=10 INITMLPERF=1 BERT_LAYERS=2 python3 examples/mlperf/model_train.py | tee $LOGFILE
+BENCHMARK=10 INITMLPERF=1 RESET_STEP=1 python3 examples/mlperf/model_train.py | tee $LOGFILE

 # run
 PARALLEL=0 RUNMLPERF=1 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -4,20 +4,24 @@ This problem uses BERT for NLP.

 ## Requirements

-Install tinygrad and mlperf-logging (uncomment mlperf from setup.py) from branch mlperf_training_v5.0.
+Install tinygrad and mlperf-logging from master.
 ```
 git clone https://github.com/tinygrad/tinygrad.git
 python3 -m pip install -e ".[mlperf]"
 ```
-Also install gdown (for dataset), numpy, tqdm and tensorflow.
+Also install tqdm and tensorflow.
 ```
-pip install gdown numpy tqdm tensorflow
+pip install tqdm tensorflow
 ```

 ### tinybox_green
 Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
 This is the default on production tinybox green.

+### tinybox_red
+Disable cwsr + increase mes timeout.
+Install the custom amdgpu driver per [README](https://github.com/nimlgen/amdgpu_ubuntu_22_04/blob/v6.1.3/readme.md)
+
 # 2. Directions

 ## Steps to download and verify data
@@ -52,18 +56,18 @@ BASEDIR="/raid/datasets/wiki" python3 extra/datasets/wikipedia.py pre-eval

 #### Steps to run benchmark
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_green/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_green/run_and_time.sh
 ```

 ### tinybox_red

-#### Steps to run benchmark
+#### One time setup
+
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_red/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_red/setup.sh
 ```
-### tinybox_8xMI300X

 #### Steps to run benchmark
 ```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/bert/implementations/tinybox_8xMI300X/run_and_time.sh
+examples/mlperf/training_submission_v4.1/tinycorp/benchmarks/bert/implementations/tinybox_red/run_and_time.sh
 ```
@@ -1,17 +1,14 @@
 #!/bin/bash

-export PYTHONPATH="." AMD=1
+export PYTHONPATH="."
 export MODEL="bert"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=5 BEAM_UOPS_MAX=10000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=3 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
-export BEAM_LOG_SURPASS_MAX=1
 export BASEDIR="/raid/datasets/wiki"

 export RESET_STEP=1
-export BENCHMARK=10 BERT_LAYERS=2 DEBUG=2
+export BENCHMARK=10 DEBUG=2

 python3 examples/mlperf/model_train.py
@@ -1,12 +1,10 @@
 #!/bin/bash

-export PYTHONPATH="." AMD=1
+export PYTHONPATH="."
 export MODEL="bert"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=5 BEAM_UOPS_MAX=10000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=3 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
 export BASEDIR="/raid/datasets/wiki"

@@ -1,13 +1,11 @@
 #!/bin/bash

-export PYTHONPATH="." AMD=1
+export PYTHONPATH="."
 export MODEL="bert"
 export SUBMISSION_PLATFORM="tinybox_red"
 export DEFAULT_FLOAT="HALF" SUM_DTYPE="HALF" GPUS=6 BS=96 EVAL_BS=96

-export FUSE_ARANGE=1 FUSE_ARANGE_UINT=0
-
-export BEAM=5 BEAM_UOPS_MAX=8000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
+export BEAM=3 BEAM_UOPS_MAX=3000 BEAM_UPCAST_MAX=256 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5
 export IGNORE_JIT_FIRST_BEAM=1
 export BASEDIR="/raid/datasets/wiki"

@@ -18,13 +16,8 @@ export SEED=$RANDOM
 DATETIME=$(date "+%m%d%H%M")
 LOGFILE="bert_red_${DATETIME}_${SEED}.log"

-export HCQDEV_WAIT_TIMEOUT_MS=100000  # prevents hang?
-
 # init
-sleep 5 && sudo rmmod amdgpu || true
-BENCHMARK=10 INITMLPERF=1 BERT_LAYERS=2 python3 examples/mlperf/model_train.py | tee $LOGFILE
+BENCHMARK=10 INITMLPERF=1 RESET_STEP=1 python3 examples/mlperf/model_train.py | tee $LOGFILE

 # run
-# TODO: AM driver resulted in nan
-sudo modprobe amdgpu
 PARALLEL=0 RUNMLPERF=1 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -0,0 +1,8 @@
+#!/bin/bash
+
+rocm-smi --setprofile compute
+rocm-smi --setmclk 3
+rocm-smi --setperflevel high
+
+# power cap to 350W
+# echo "350000000" | sudo tee /sys/class/drm/card{1..6}/device/hwmon/hwmon*/power1_cap
@@ -0,0 +1,50 @@
+# 1. Problem
+
+This problem uses the ResNet-50 CNN to do image classification.
+
+## Requirements
+
+Install tinygrad and mlperf-logging from master.
+```
+git clone https://github.com/tinygrad/tinygrad.git
+python3 -m pip install -e ".[mlperf]"
+```
+
+### tinybox_green
+Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
+This is the default on production tinybox green.
+
+### tinybox_red
+Disable cwsr
+This is the default on production tinybox red.
+```
+sudo vi /etc/modprobe.d/amdgpu.conf
+cat <<EOF > /etc/modprobe.d/amdgpu.conf
+options amdgpu cwsr_enable=0
+EOF
+sudo update-initramfs -u
+sudo reboot
+
+# validate
+sudo cat /sys/module/amdgpu/parameters/cwsr_enable #= 0
+```
+
+# 2. Directions
+
+## Steps to download and verify data
+
+```
+IMGNET_TRAIN=1 python3 extra/datasets/imagenet_download.py
+```
+
+## Steps for one time setup
+
+### tinybox_red
+```
+examples/mlperf/training_submission_v4.0/tinycorp/benchmarks/resnet/implementations/tinybox_red/setup.sh
+```
+
+## Steps to run benchmark
+```
+examples/mlperf/training_submission_v4.0/tinycorp/benchmarks/resnet/implementations/tinybox_red/run_and_time.sh
+```
@@ -0,0 +1,13 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=10 BEAM_PADTO=0
+
+export BENCHMARK=10 DEBUG=2
+
+python3 examples/mlperf/model_train.py
@@ -0,0 +1,15 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=10 BEAM_PADTO=0
+
+export EVAL_START_EPOCH=3 EVAL_FREQ=4
+
+export WANDB=1 PARALLEL=0
+
+python3 examples/mlperf/model_train.py
@@ -0,0 +1,23 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export SUBMISSION_PLATFORM="tinybox_green"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=10 BEAM_PADTO=0
+
+# pip install -e ".[mlperf]"
+export LOGMLPERF=1
+
+export SEED=$RANDOM
+DATETIME=$(date "+%m%d%H%M")
+LOGFILE="resnet_green_${DATETIME}_${SEED}.log"
+
+# init
+BENCHMARK=10 INITMLPERF=1 python3 examples/mlperf/model_train.py | tee $LOGFILE
+
+# run
+PARALLEL=0 RUNMLPERF=1 EVAL_START_EPOCH=3 EVAL_FREQ=4 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -0,0 +1,50 @@
+# 1. Problem
+
+This problem uses the ResNet-50 CNN to do image classification.
+
+## Requirements
+
+Install tinygrad and mlperf-logging from master.
+```
+git clone https://github.com/tinygrad/tinygrad.git
+python3 -m pip install -e ".[mlperf]"
+```
+
+### tinybox_green
+Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
+This is the default on production tinybox green.
+
+### tinybox_red
+Disable cwsr
+This is the default on production tinybox red.
+```
+sudo vi /etc/modprobe.d/amdgpu.conf
+cat <<EOF > /etc/modprobe.d/amdgpu.conf
+options amdgpu cwsr_enable=0
+EOF
+sudo update-initramfs -u
+sudo reboot
+
+# validate
+sudo cat /sys/module/amdgpu/parameters/cwsr_enable #= 0
+```
+
+# 2. Directions
+
+## Steps to download and verify data
+
+```
+IMGNET_TRAIN=1 python3 extra/datasets/imagenet_download.py
+```
+
+## Steps for one time setup
+
+### tinybox_red
+```
+examples/mlperf/training_submission_v4.0/tinycorp/benchmarks/resnet/implementations/tinybox_red/setup.sh
+```
+
+## Steps to run benchmark
+```
+examples/mlperf/training_submission_v4.0/tinycorp/benchmarks/resnet/implementations/tinybox_red/run_and_time.sh
+```
@@ -0,0 +1,13 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=2000 BEAM_UPCAST_MAX=96 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
+
+export BENCHMARK=10 DEBUG=2
+
+python3 examples/mlperf/model_train.py
@@ -0,0 +1,15 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=2000 BEAM_UPCAST_MAX=96 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
+
+export EVAL_START_EPOCH=3 EVAL_FREQ=4
+
+export WANDB=1 PARALLEL=0
+
+python3 examples/mlperf/model_train.py
@@ -0,0 +1,23 @@
+#!/bin/bash
+
+export PYTHONPATH="."
+export MODEL="resnet"
+export SUBMISSION_PLATFORM="tinybox_red"
+export DEFAULT_FLOAT="HALF" GPUS=6 BS=1536 EVAL_BS=192
+
+export RESET_STEP=0
+
+export TRAIN_BEAM=4 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=2000 BEAM_UPCAST_MAX=96 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
+
+# pip install -e ".[mlperf]"
+export LOGMLPERF=1
+
+export SEED=$RANDOM
+DATETIME=$(date "+%m%d%H%M")
+LOGFILE="resnet_red_${DATETIME}_${SEED}.log"
+
+# init
+BENCHMARK=10 INITMLPERF=1 python3 examples/mlperf/model_train.py | tee $LOGFILE
+
+# run
+PARALLEL=0 RUNMLPERF=1 EVAL_START_EPOCH=3 EVAL_FREQ=4 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -0,0 +1,8 @@
+#!/bin/bash
+
+rocm-smi --setprofile compute
+rocm-smi --setmclk 3
+rocm-smi --setperflevel high
+
+# power cap to 350W
+echo "350000000" | sudo tee /sys/class/drm/card{1..6}/device/hwmon/hwmon*/power1_cap
@@ -1,38 +0,0 @@
-# 1. Problem
-
-This problem uses RetinaNet for SSD.
-
-## Requirements
-
-Install tinygrad and mlperf-logging (uncomment mlperf from setup.py) from branch mlperf_training_v5.0.
-```
-git clone https://github.com/tinygrad/tinygrad.git
-python3 -m pip install -e ".[mlperf]"
-```
-
-Also install the following dependencies:
-```
-pip install tqdm numpy pycocotools boto3 pandas torch torchvision
-```
-
-### tinybox_green
-Install the p2p driver per [README](https://github.com/tinygrad/open-gpu-kernel-modules/blob/550.54.15-p2p/README.md)
-This is the default on production tinybox green.
-
-# 2. Directions
-
-## Steps to download data
-
-Run the following:
-```
-BASEDIR=/raid/datasets/openimages python3 extra/datasets/openimages.py
-```
-
-## Running
-
-### tinybox_green
-
-#### Steps to run benchmark
-```
-examples/mlperf/training_submission_v5.0/tinycorp/benchmarks/retinanet/implementations/tinybox_green/run_and_time.sh
-```
@@ -1,14 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." NV=1
-export MODEL="retinanet"
-export DEFAULT_FLOAT="HALF" GPUS=6 BS=96 EVAL_BS=96
-export BASEDIR="/raid/datasets/openimages"
-
-# export RESET_STEP=0
-
-export TRAIN_BEAM=2 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
-
-export BENCHMARK=5 DEBUG=2
-
-python examples/mlperf/model_train.py
@@ -1,15 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." NV=1
-export MODEL="retinanet"
-export DEFAULT_FLOAT="HALF" GPUS=6 BS=96 EVAL_BS=96
-export BASEDIR="/raid/datasets/openimages"
-
-# export RESET_STEP=0
-
-export TRAIN_BEAM=2 IGNORE_JIT_FIRST_BEAM=1 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
-
-export WANDB=1 PARALLEL=0
-export RUNMLPERF=1
-
-python examples/mlperf/model_train.py
@@ -1,23 +0,0 @@
-#!/bin/bash
-
-export PYTHONPATH="." NV=1
-export MODEL="retinanet"
-export SUBMISSION_PLATFORM="tinybox_green"
-export DEFAULT_FLOAT="HALF" GPUS=6 BS=96 EVAL_BS=96
-
-export TRAIN_BEAM=2 BEAM_UOPS_MAX=1500 BEAM_UPCAST_MAX=64 BEAM_LOCAL_MAX=1024 BEAM_MIN_PROGRESS=5 BEAM_PADTO=0
-export IGNORE_JIT_FIRST_BEAM=1
-export BASEDIR="/raid/datasets/openimages"
-
-# pip install -e ".[mlperf]"
-export LOGMLPERF=1
-
-export SEED=$RANDOM
-DATETIME=$(date "+%m%d%H%M")
-LOGFILE="retinanet_green_${DATETIME}_${SEED}.log"
-
-# init
-BENCHMARK=10 INITMLPERF=1 python3 examples/mlperf/model_train.py | tee $LOGFILE
-
-# run
-PARALLEL=0 RUNMLPERF=1 python3 examples/mlperf/model_train.py | tee -a $LOGFILE
@@ -1,96 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745530925055, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925071, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925071, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925071, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925071, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925111, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745530925112, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745532610164, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745532618573, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745532618590, "event_type": "POINT_IN_TIME", "key": "seed", "value": 28597, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631107, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631108, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745532631109, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745532663595, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745533321994, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745533363168, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745533363168, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37238641977310183, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.37238641977310183}}
-:::MLLOG {"namespace": "", "time_ms": 1745533638843, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745533643497, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745533643497, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3876396149396896, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.3876396149396896}}
-:::MLLOG {"namespace": "", "time_ms": 1745533919587, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745533924272, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745533924272, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4356461137533188, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.4356461137533188}}
-:::MLLOG {"namespace": "", "time_ms": 1745534200765, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745534205421, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745534205422, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.523741614818573, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.523741614818573}}
-:::MLLOG {"namespace": "", "time_ms": 1745534481862, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745534486501, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745534486501, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6327109396457672, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6327109396457672}}
-:::MLLOG {"namespace": "", "time_ms": 1745534762710, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745534767349, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745534767349, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6953712999820709, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.6953712999820709}}
-:::MLLOG {"namespace": "", "time_ms": 1745535043273, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745535047907, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745535047907, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7062251627445221, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7062251627445221}}
-:::MLLOG {"namespace": "", "time_ms": 1745535324088, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745535328759, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745535328760, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7097240626811981, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7097240626811981}}
-:::MLLOG {"namespace": "", "time_ms": 1745535604822, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745535609479, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745535609479, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7112975955009461, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7112975955009461}}
-:::MLLOG {"namespace": "", "time_ms": 1745535885084, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745535889738, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745535889738, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.711781257390976, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.711781257390976}}
-:::MLLOG {"namespace": "", "time_ms": 1745536165354, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745536170008, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745536170009, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7138555943965912, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7138555943965912}}
-:::MLLOG {"namespace": "", "time_ms": 1745536445570, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745536450183, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745536450183, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7145828008651733, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7145828008651733}}
-:::MLLOG {"namespace": "", "time_ms": 1745536725721, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745536730380, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745536730381, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7148040235042572, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7148040235042572}}
-:::MLLOG {"namespace": "", "time_ms": 1745537006570, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745537011215, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745537011215, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7158548653125762, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7158548653125762}}
-:::MLLOG {"namespace": "", "time_ms": 1745537289257, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745537293941, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745537293941, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160437941551209, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7160437941551209}}
-:::MLLOG {"namespace": "", "time_ms": 1745537580273, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745537584925, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745537584926, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7169912159442902, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7169912159442902}}
-:::MLLOG {"namespace": "", "time_ms": 1745537860132, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745537864770, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745537864770, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7176614046096802, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7176614046096802}}
-:::MLLOG {"namespace": "", "time_ms": 1745538142023, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745538146643, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745538146644, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182221412658691, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7182221412658691}}
-:::MLLOG {"namespace": "", "time_ms": 1745538425213, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745538429916, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745538429916, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186778724193573, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7186778724193573}}
-:::MLLOG {"namespace": "", "time_ms": 1745538705412, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745538710041, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745538710041, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190635979175568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7190635979175568}}
-:::MLLOG {"namespace": "", "time_ms": 1745538992202, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745538996846, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745538996846, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7196638345718384, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.7196638345718384}}
-:::MLLOG {"namespace": "", "time_ms": 1745539272384, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745539277037, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745539277037, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7201055765151978, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.7201055765151978}}
-:::MLLOG {"namespace": "", "time_ms": 1745539277038, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3289088}}
-:::MLLOG {"namespace": "", "time_ms": 1745539277038, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,87 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745539327854, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327869, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327870, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327870, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327870, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327912, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745539327912, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745540911909, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745540920744, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745540920760, "event_type": "POINT_IN_TIME", "key": "seed", "value": 15393, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933147, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933148, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745540933149, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745540962253, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745541620433, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745541661192, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745541661193, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37288502752780917, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.37288502752780917}}
-:::MLLOG {"namespace": "", "time_ms": 1745541940327, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745541944936, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745541944936, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3860586792230606, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.3860586792230606}}
-:::MLLOG {"namespace": "", "time_ms": 1745542224659, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745542229287, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745542229287, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.429127961397171, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.429127961397171}}
-:::MLLOG {"namespace": "", "time_ms": 1745542510137, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745542514771, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745542514771, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.529323160648346, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.529323160648346}}
-:::MLLOG {"namespace": "", "time_ms": 1745542796386, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745542801034, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745542801034, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6631237685680389, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6631237685680389}}
-:::MLLOG {"namespace": "", "time_ms": 1745543082494, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745543087128, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745543087128, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7024946749210358, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.7024946749210358}}
-:::MLLOG {"namespace": "", "time_ms": 1745543367964, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745543372586, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745543372587, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.708392471075058, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.708392471075058}}
-:::MLLOG {"namespace": "", "time_ms": 1745543653532, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745543658174, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745543658174, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.711252635717392, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.711252635717392}}
-:::MLLOG {"namespace": "", "time_ms": 1745543939454, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745543944107, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745543944107, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7130857110023499, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7130857110023499}}
-:::MLLOG {"namespace": "", "time_ms": 1745544225356, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745544230002, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745544230002, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7136700868606567, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7136700868606567}}
-:::MLLOG {"namespace": "", "time_ms": 1745544512872, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745544517504, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745544517504, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7149688005447388, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7149688005447388}}
-:::MLLOG {"namespace": "", "time_ms": 1745544801071, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745544805699, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745544805700, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.715419614315033, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.715419614315033}}
-:::MLLOG {"namespace": "", "time_ms": 1745545085900, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745545090549, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745545090549, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160871505737305, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7160871505737305}}
-:::MLLOG {"namespace": "", "time_ms": 1745545375103, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745545379758, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745545379759, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7170936107635498, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7170936107635498}}
-:::MLLOG {"namespace": "", "time_ms": 1745545660444, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745545665097, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745545665097, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717720341682434, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.717720341682434}}
-:::MLLOG {"namespace": "", "time_ms": 1745545946029, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745545950685, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745545950686, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7183034479618072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7183034479618072}}
-:::MLLOG {"namespace": "", "time_ms": 1745546233082, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745546237727, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745546237727, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182737410068512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7182737410068512}}
-:::MLLOG {"namespace": "", "time_ms": 1745546517916, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745546522570, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745546522571, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191624820232392, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7191624820232392}}
-:::MLLOG {"namespace": "", "time_ms": 1745546806258, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745546810904, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745546810904, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200840294361115, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7200840294361115}}
-:::MLLOG {"namespace": "", "time_ms": 1745546810904, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2840576}}
-:::MLLOG {"namespace": "", "time_ms": 1745546810905, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745546859441, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859457, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859457, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859457, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859457, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859498, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745546859498, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745548472183, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745548480461, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745548480478, "event_type": "POINT_IN_TIME", "key": "seed", "value": 16206, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492724, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492725, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745548492726, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745548524140, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745549187391, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745549229147, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745549229147, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3727019250392914, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.3727019250392914}}
-:::MLLOG {"namespace": "", "time_ms": 1745549513996, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745549518578, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745549518578, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3877880424261093, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.3877880424261093}}
-:::MLLOG {"namespace": "", "time_ms": 1745549804473, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745549809036, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745549809036, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4411086171865463, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.4411086171865463}}
-:::MLLOG {"namespace": "", "time_ms": 1745550093860, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745550098426, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745550098426, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5526047289371491, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5526047289371491}}
-:::MLLOG {"namespace": "", "time_ms": 1745550383034, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745550387588, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745550387589, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6731186151504517, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6731186151504517}}
-:::MLLOG {"namespace": "", "time_ms": 1745550672196, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745550676767, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745550676767, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7031940758228302, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.7031940758228302}}
-:::MLLOG {"namespace": "", "time_ms": 1745550961962, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745550966515, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745550966515, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7088134288787842, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7088134288787842}}
-:::MLLOG {"namespace": "", "time_ms": 1745551251075, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745551255650, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745551255650, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7112709581851959, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7112709581851959}}
-:::MLLOG {"namespace": "", "time_ms": 1745551539957, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745551544529, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745551544530, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7128494203090667, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7128494203090667}}
-:::MLLOG {"namespace": "", "time_ms": 1745551828589, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745551833167, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745551833167, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135251462459564, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7135251462459564}}
-:::MLLOG {"namespace": "", "time_ms": 1745552120068, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745552124628, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745552124628, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147472620010376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7147472620010376}}
-:::MLLOG {"namespace": "", "time_ms": 1745552408212, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745552412794, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745552412794, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7153081774711609, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7153081774711609}}
-:::MLLOG {"namespace": "", "time_ms": 1745552697710, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745552702247, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745552702247, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163431167602539, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7163431167602539}}
-:::MLLOG {"namespace": "", "time_ms": 1745552985842, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745552990399, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745552990399, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7166160047054291, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7166160047054291}}
-:::MLLOG {"namespace": "", "time_ms": 1745553277481, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745553282034, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745553282035, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172852098941803, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7172852098941803}}
-:::MLLOG {"namespace": "", "time_ms": 1745553566684, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745553571258, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745553571259, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7169495701789856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7169495701789856}}
-:::MLLOG {"namespace": "", "time_ms": 1745553854981, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745553859558, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745553859559, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182259142398835, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7182259142398835}}
-:::MLLOG {"namespace": "", "time_ms": 1745554143398, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745554147970, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745554147970, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718964672088623, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.718964672088623}}
-:::MLLOG {"namespace": "", "time_ms": 1745554435137, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745554439700, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745554439700, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7193355560302734, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7193355560302734}}
-:::MLLOG {"namespace": "", "time_ms": 1745554725252, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745554729828, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745554729828, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7201014399528504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7201014399528504}}
-:::MLLOG {"namespace": "", "time_ms": 1745554729829, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2990080}}
-:::MLLOG {"namespace": "", "time_ms": 1745554729829, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745554777762, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777778, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777778, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777778, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777778, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777819, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745554777819, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745556404094, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745556412199, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745556412215, "event_type": "POINT_IN_TIME", "key": "seed", "value": 20163, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424820, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424820, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424821, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745556424822, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745556457286, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745557115312, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745557156018, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745557156018, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3721550852060318, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.3721550852060318}}
-:::MLLOG {"namespace": "", "time_ms": 1745557435176, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745557439784, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745557439784, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3927359789609909, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.3927359789609909}}
-:::MLLOG {"namespace": "", "time_ms": 1745557719263, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745557723910, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745557723910, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.44803847670555114, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.44803847670555114}}
-:::MLLOG {"namespace": "", "time_ms": 1745558003706, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745558008350, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745558008350, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5591910660266877, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5591910660266877}}
-:::MLLOG {"namespace": "", "time_ms": 1745558288347, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745558292987, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745558292987, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6610698461532593, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6610698461532593}}
-:::MLLOG {"namespace": "", "time_ms": 1745558573332, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745558578000, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745558578000, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7006066799163818, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.7006066799163818}}
-:::MLLOG {"namespace": "", "time_ms": 1745558858774, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745558863431, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745558863431, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7073053598403931, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7073053598403931}}
-:::MLLOG {"namespace": "", "time_ms": 1745559143329, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745559147961, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745559147962, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7103267848491669, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7103267848491669}}
-:::MLLOG {"namespace": "", "time_ms": 1745559427612, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745559432260, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745559432260, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7116473376750946, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7116473376750946}}
-:::MLLOG {"namespace": "", "time_ms": 1745559711261, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745559715880, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745559715881, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7130024552345275, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7130024552345275}}
-:::MLLOG {"namespace": "", "time_ms": 1745559994602, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745559999235, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745559999235, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7140777409076691, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7140777409076691}}
-:::MLLOG {"namespace": "", "time_ms": 1745560278151, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745560282795, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745560282796, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.714939397573471, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.714939397573471}}
-:::MLLOG {"namespace": "", "time_ms": 1745560561092, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745560565689, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745560565690, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7154468119144439, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7154468119144439}}
-:::MLLOG {"namespace": "", "time_ms": 1745560849874, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745560854474, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745560854474, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7161401331424713, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7161401331424713}}
-:::MLLOG {"namespace": "", "time_ms": 1745561132673, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745561137319, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745561137319, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163729786872863, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7163729786872863}}
-:::MLLOG {"namespace": "", "time_ms": 1745561418781, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745561423392, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745561423393, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717225980758667, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.717225980758667}}
-:::MLLOG {"namespace": "", "time_ms": 1745561701170, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745561705798, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745561705798, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181021451950074, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7181021451950074}}
-:::MLLOG {"namespace": "", "time_ms": 1745561988084, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745561992722, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745561992722, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191600739955902, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7191600739955902}}
-:::MLLOG {"namespace": "", "time_ms": 1745562273977, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745562278658, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745562278659, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719371110200882, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.719371110200882}}
-:::MLLOG {"namespace": "", "time_ms": 1745562559865, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745562564486, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745562564486, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718993628025055, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.718993628025055}}
-:::MLLOG {"namespace": "", "time_ms": 1745562845123, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745562849755, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745562849756, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200249254703521, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.7200249254703521}}
-:::MLLOG {"namespace": "", "time_ms": 1745562849756, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3139584}}
-:::MLLOG {"namespace": "", "time_ms": 1745562849756, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,96 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745562899619, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899635, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899635, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899635, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899635, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899676, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745562899676, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745564587425, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745564596874, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745564596890, "event_type": "POINT_IN_TIME", "key": "seed", "value": 13352, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609291, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609292, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745564609293, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745564642816, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745565297838, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745565339344, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745565339345, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3725235253572464, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.3725235253572464}}
-:::MLLOG {"namespace": "", "time_ms": 1745565617369, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745565622024, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745565622025, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.38939482867717745, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.38939482867717745}}
-:::MLLOG {"namespace": "", "time_ms": 1745565900583, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745565905232, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745565905233, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.43388367593288424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.43388367593288424}}
-:::MLLOG {"namespace": "", "time_ms": 1745566183301, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745566187916, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745566187917, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5344274997711181, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5344274997711181}}
-:::MLLOG {"namespace": "", "time_ms": 1745566466251, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745566470898, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745566470898, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6529988288879395, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6529988288879395}}
-:::MLLOG {"namespace": "", "time_ms": 1745566749556, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745566754223, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745566754223, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6987890124320983, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.6987890124320983}}
-:::MLLOG {"namespace": "", "time_ms": 1745567033420, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745567038073, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745567038073, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7056617558002471, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7056617558002471}}
-:::MLLOG {"namespace": "", "time_ms": 1745567316228, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745567320871, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745567320871, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7096020996570587, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7096020996570587}}
-:::MLLOG {"namespace": "", "time_ms": 1745567599127, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745567603785, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745567603785, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7113405406475067, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7113405406475067}}
-:::MLLOG {"namespace": "", "time_ms": 1745567881894, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745567886538, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745567886539, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7122747898101807, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7122747898101807}}
-:::MLLOG {"namespace": "", "time_ms": 1745568164600, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745568169238, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745568169239, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7133678615093231, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7133678615093231}}
-:::MLLOG {"namespace": "", "time_ms": 1745568447131, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745568451771, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745568451771, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7142562866210938, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7142562866210938}}
-:::MLLOG {"namespace": "", "time_ms": 1745568731625, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745568736289, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745568736290, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7150763928890228, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7150763928890228}}
-:::MLLOG {"namespace": "", "time_ms": 1745569015501, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745569020163, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745569020163, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7155498623847961, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7155498623847961}}
-:::MLLOG {"namespace": "", "time_ms": 1745569298438, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745569303066, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745569303066, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160784900188446, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7160784900188446}}
-:::MLLOG {"namespace": "", "time_ms": 1745569581278, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745569585933, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745569585933, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7170640766620636, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7170640766620636}}
-:::MLLOG {"namespace": "", "time_ms": 1745569866592, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745569871241, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745569871241, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7173652410507202, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7173652410507202}}
-:::MLLOG {"namespace": "", "time_ms": 1745570152468, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745570157114, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745570157115, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718361359834671, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.718361359834671}}
-:::MLLOG {"namespace": "", "time_ms": 1745570435938, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745570440644, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745570440644, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186611413955688, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7186611413955688}}
-:::MLLOG {"namespace": "", "time_ms": 1745570726442, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745570731082, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745570731082, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191152453422547, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7191152453422547}}
-:::MLLOG {"namespace": "", "time_ms": 1745571012781, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745571017427, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745571017427, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7195845186710358, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.7195845186710358}}
-:::MLLOG {"namespace": "", "time_ms": 1745571300015, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745571304681, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745571304682, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200073778629303, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.7200073778629303}}
-:::MLLOG {"namespace": "", "time_ms": 1745571304682, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3289088}}
-:::MLLOG {"namespace": "", "time_ms": 1745571304682, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,96 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745571354572, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354587, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354587, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354587, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354587, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354629, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745571354629, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745573082356, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745573090559, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745573090575, "event_type": "POINT_IN_TIME", "key": "seed", "value": 29862, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103073, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103073, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103073, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103074, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103075, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103075, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103075, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745573103075, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745573134049, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745573811088, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745573852243, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745573852243, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3729754567146301, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.3729754567146301}}
-:::MLLOG {"namespace": "", "time_ms": 1745574135466, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745574140258, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745574140258, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.38697501420974734, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.38697501420974734}}
-:::MLLOG {"namespace": "", "time_ms": 1745574423983, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745574428791, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745574428791, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.42467189133167266, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.42467189133167266}}
-:::MLLOG {"namespace": "", "time_ms": 1745574712426, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745574717211, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745574717212, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5411350011825562, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5411350011825562}}
-:::MLLOG {"namespace": "", "time_ms": 1745575000437, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745575005233, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745575005233, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6546417593955993, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6546417593955993}}
-:::MLLOG {"namespace": "", "time_ms": 1745575287460, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745575292227, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745575292227, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7013923466205597, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.7013923466205597}}
-:::MLLOG {"namespace": "", "time_ms": 1745575574252, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745575579038, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745575579038, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7082416176795959, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7082416176795959}}
-:::MLLOG {"namespace": "", "time_ms": 1745575860880, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745575865649, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745575865649, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7109045207500457, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7109045207500457}}
-:::MLLOG {"namespace": "", "time_ms": 1745576147355, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745576152107, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745576152108, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7121940672397613, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7121940672397613}}
-:::MLLOG {"namespace": "", "time_ms": 1745576433426, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745576438176, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745576438176, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.713076388835907, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.713076388835907}}
-:::MLLOG {"namespace": "", "time_ms": 1745576719437, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745576724186, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745576724186, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147435486316681, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7147435486316681}}
-:::MLLOG {"namespace": "", "time_ms": 1745577005513, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745577010268, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745577010268, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152598202228546, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7152598202228546}}
-:::MLLOG {"namespace": "", "time_ms": 1745577294417, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745577299176, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745577299177, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7158841907978057, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7158841907978057}}
-:::MLLOG {"namespace": "", "time_ms": 1745577583196, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745577587938, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745577587938, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162841558456421, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7162841558456421}}
-:::MLLOG {"namespace": "", "time_ms": 1745577872125, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745577876872, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745577876872, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7171514630317688, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7171514630317688}}
-:::MLLOG {"namespace": "", "time_ms": 1745578157392, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745578162167, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745578162167, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717557144165039, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.717557144165039}}
-:::MLLOG {"namespace": "", "time_ms": 1745578442695, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745578447477, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745578447478, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181805670261383, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7181805670261383}}
-:::MLLOG {"namespace": "", "time_ms": 1745578732198, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745578736960, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745578736960, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7193208992481231, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7193208992481231}}
-:::MLLOG {"namespace": "", "time_ms": 1745579017075, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745579021830, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745579021831, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190930187702179, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7190930187702179}}
-:::MLLOG {"namespace": "", "time_ms": 1745579308467, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745579313199, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745579313199, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7193685650825501, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7193685650825501}}
-:::MLLOG {"namespace": "", "time_ms": 1745579595974, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745579600728, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745579600728, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198175132274628, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.7198175132274628}}
-:::MLLOG {"namespace": "", "time_ms": 1745579881098, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745579885849, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745579885849, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.720059609413147, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.720059609413147}}
-:::MLLOG {"namespace": "", "time_ms": 1745579885849, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3289088}}
-:::MLLOG {"namespace": "", "time_ms": 1745579885849, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,108 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745579935971, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745579935986, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745579935987, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745579935987, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745579935987, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745579936029, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745579936029, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745581719259, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745581728446, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745581728462, "event_type": "POINT_IN_TIME", "key": "seed", "value": 7754, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740992, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740993, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745581740994, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745581773454, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745582423629, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745582464633, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745582464633, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37272038757801057, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.37272038757801057}}
-:::MLLOG {"namespace": "", "time_ms": 1745582740508, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745582745067, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745582745068, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.38925057649612427, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.38925057649612427}}
-:::MLLOG {"namespace": "", "time_ms": 1745583021215, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745583025776, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745583025776, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.43427990674972533, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.43427990674972533}}
-:::MLLOG {"namespace": "", "time_ms": 1745583301981, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745583306560, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745583306561, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5278923571109772, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5278923571109772}}
-:::MLLOG {"namespace": "", "time_ms": 1745583582767, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745583587345, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745583587345, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6390926957130432, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6390926957130432}}
-:::MLLOG {"namespace": "", "time_ms": 1745583863550, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745583868135, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745583868136, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6956972718238831, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.6956972718238831}}
-:::MLLOG {"namespace": "", "time_ms": 1745584144322, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745584148929, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745584148929, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.705748564004898, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.705748564004898}}
-:::MLLOG {"namespace": "", "time_ms": 1745584425116, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745584429704, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745584429704, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7098638355731964, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7098638355731964}}
-:::MLLOG {"namespace": "", "time_ms": 1745584705931, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745584710512, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745584710512, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7110268771648407, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7110268771648407}}
-:::MLLOG {"namespace": "", "time_ms": 1745584986656, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745584991214, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745584991214, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.712305212020874, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.712305212020874}}
-:::MLLOG {"namespace": "", "time_ms": 1745585267330, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745585271920, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745585271921, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135731339454651, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7135731339454651}}
-:::MLLOG {"namespace": "", "time_ms": 1745585551669, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745585556247, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745585556248, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7142563343048096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7142563343048096}}
-:::MLLOG {"namespace": "", "time_ms": 1745585833849, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745585838410, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745585838410, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152255415916443, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7152255415916443}}
-:::MLLOG {"namespace": "", "time_ms": 1745586114185, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745586118755, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745586118755, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7159430027008057, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7159430027008057}}
-:::MLLOG {"namespace": "", "time_ms": 1745586394477, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745586399048, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745586399048, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162509083747863, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7162509083747863}}
-:::MLLOG {"namespace": "", "time_ms": 1745586674813, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745586679413, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745586679414, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7167587816715241, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7167587816715241}}
-:::MLLOG {"namespace": "", "time_ms": 1745586958455, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745586963029, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745586963029, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172225117683411, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7172225117683411}}
-:::MLLOG {"namespace": "", "time_ms": 1745587244889, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745587249485, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745587249485, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7183384358882904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7183384358882904}}
-:::MLLOG {"namespace": "", "time_ms": 1745587524929, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745587529566, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745587529566, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7180015325546265, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7180015325546265}}
-:::MLLOG {"namespace": "", "time_ms": 1745587808591, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745587813157, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745587813158, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7184844195842743, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7184844195842743}}
-:::MLLOG {"namespace": "", "time_ms": 1745588093001, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745588097586, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745588097587, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718977439403534, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.718977439403534}}
-:::MLLOG {"namespace": "", "time_ms": 1745588375341, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745588379922, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745588379922, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192779302597045, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.7192779302597045}}
-:::MLLOG {"namespace": "", "time_ms": 1745588655781, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3438592, "step_num": 3358}}
-:::MLLOG {"namespace": "", "time_ms": 1745588660364, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3438592, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3438592, "step_num": 3358, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745588660364, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719474196434021, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3438592, "masked_lm_accuracy": 0.719474196434021}}
-:::MLLOG {"namespace": "", "time_ms": 1745588936187, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3588096, "step_num": 3504}}
-:::MLLOG {"namespace": "", "time_ms": 1745588940778, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3588096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3588096, "step_num": 3504, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745588940778, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198640763759613, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3588096, "masked_lm_accuracy": 0.7198640763759613}}
-:::MLLOG {"namespace": "", "time_ms": 1745589216640, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3737600, "step_num": 3650}}
-:::MLLOG {"namespace": "", "time_ms": 1745589221228, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3737600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3737600, "step_num": 3650, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745589221228, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7199562847614288, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3737600, "masked_lm_accuracy": 0.7199562847614288}}
-:::MLLOG {"namespace": "", "time_ms": 1745589497055, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3887104, "step_num": 3796}}
-:::MLLOG {"namespace": "", "time_ms": 1745589501645, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3887104, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3887104, "step_num": 3796, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745589501646, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7206668496131897, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3887104, "masked_lm_accuracy": 0.7206668496131897}}
-:::MLLOG {"namespace": "", "time_ms": 1745589501696, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3887104, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3887104}}
-:::MLLOG {"namespace": "", "time_ms": 1745589501696, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,105 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745589550561, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550577, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550577, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550577, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550577, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550621, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745589550622, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745591179400, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745591187432, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745591187448, "event_type": "POINT_IN_TIME", "key": "seed", "value": 25385, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199679, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199680, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745591199681, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745591230921, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745591901120, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745591941186, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745591941187, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37098502218723295, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.37098502218723295}}
-:::MLLOG {"namespace": "", "time_ms": 1745592221446, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745592226014, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745592226015, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.38992435932159425, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.38992435932159425}}
-:::MLLOG {"namespace": "", "time_ms": 1745592506825, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745592511386, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745592511386, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4360688954591751, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.4360688954591751}}
-:::MLLOG {"namespace": "", "time_ms": 1745592792597, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745592797166, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745592797167, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5389649093151092, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5389649093151092}}
-:::MLLOG {"namespace": "", "time_ms": 1745593078507, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745593083063, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745593083064, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6623023927211762, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6623023927211762}}
-:::MLLOG {"namespace": "", "time_ms": 1745593364383, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745593368960, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745593368960, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6989677131175995, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.6989677131175995}}
-:::MLLOG {"namespace": "", "time_ms": 1745593650221, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745593654771, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745593654771, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7055331587791442, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7055331587791442}}
-:::MLLOG {"namespace": "", "time_ms": 1745593935981, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745593940552, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745593940552, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7088817238807679, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7088817238807679}}
-:::MLLOG {"namespace": "", "time_ms": 1745594221979, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745594226554, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745594226554, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7105159401893616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7105159401893616}}
-:::MLLOG {"namespace": "", "time_ms": 1745594507867, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745594512457, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745594512457, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7124811410903931, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7124811410903931}}
-:::MLLOG {"namespace": "", "time_ms": 1745594793979, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745594798544, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745594798545, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7134620428085328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7134620428085328}}
-:::MLLOG {"namespace": "", "time_ms": 1745595079851, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745595084445, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745595084445, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7141174793243408, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7141174793243408}}
-:::MLLOG {"namespace": "", "time_ms": 1745595368793, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745595373357, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745595373357, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7149844408035279, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7149844408035279}}
-:::MLLOG {"namespace": "", "time_ms": 1745595656391, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745595660943, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745595660944, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7157506585121155, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7157506585121155}}
-:::MLLOG {"namespace": "", "time_ms": 1745595946346, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745595950902, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745595950902, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160208821296692, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7160208821296692}}
-:::MLLOG {"namespace": "", "time_ms": 1745596234396, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745596238968, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745596238968, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7169745147228241, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7169745147228241}}
-:::MLLOG {"namespace": "", "time_ms": 1745596519214, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745596523783, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745596523784, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7170299768447876, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7170299768447876}}
-:::MLLOG {"namespace": "", "time_ms": 1745596804022, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745596808588, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745596808589, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186152815818787, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7186152815818787}}
-:::MLLOG {"namespace": "", "time_ms": 1745597088735, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745597093297, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745597093297, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7184043228626251, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7184043228626251}}
-:::MLLOG {"namespace": "", "time_ms": 1745597376858, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745597381430, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745597381430, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186801016330719, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7186801016330719}}
-:::MLLOG {"namespace": "", "time_ms": 1745597661697, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745597666257, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745597666257, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191139340400696, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.7191139340400696}}
-:::MLLOG {"namespace": "", "time_ms": 1745597949669, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745597954226, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745597954226, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192700445652008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.7192700445652008}}
-:::MLLOG {"namespace": "", "time_ms": 1745598236392, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3438592, "step_num": 3358}}
-:::MLLOG {"namespace": "", "time_ms": 1745598240947, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3438592, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3438592, "step_num": 3358, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745598240948, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7196909546852112, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3438592, "masked_lm_accuracy": 0.7196909546852112}}
-:::MLLOG {"namespace": "", "time_ms": 1745598521853, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3588096, "step_num": 3504}}
-:::MLLOG {"namespace": "", "time_ms": 1745598526430, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3588096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3588096, "step_num": 3504, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745598526431, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7199203789234161, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3588096, "masked_lm_accuracy": 0.7199203789234161}}
-:::MLLOG {"namespace": "", "time_ms": 1745598807367, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3737600, "step_num": 3650}}
-:::MLLOG {"namespace": "", "time_ms": 1745598811951, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3737600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3737600, "step_num": 3650, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745598811951, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200918376445771, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3737600, "masked_lm_accuracy": 0.7200918376445771}}
-:::MLLOG {"namespace": "", "time_ms": 1745598811952, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3737600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3737600}}
-:::MLLOG {"namespace": "", "time_ms": 1745598811952, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,102 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745598860595, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860610, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860610, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860610, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860610, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860653, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745598860653, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745600478970, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745600487253, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745600487271, "event_type": "POINT_IN_TIME", "key": "seed", "value": 32629, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500186, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500186, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500186, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500186, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500187, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500188, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745600500188, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745600532785, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745601205060, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745601247392, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745601247392, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3731185048818588, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.3731185048818588}}
-:::MLLOG {"namespace": "", "time_ms": 1745601525682, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745601530299, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745601530300, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.39272683262825014, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.39272683262825014}}
-:::MLLOG {"namespace": "", "time_ms": 1745601809277, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745601813902, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745601813902, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.43582180738449094, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.43582180738449094}}
-:::MLLOG {"namespace": "", "time_ms": 1745602092985, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745602097585, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745602097585, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5279603898525238, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.5279603898525238}}
-:::MLLOG {"namespace": "", "time_ms": 1745602376585, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745602381222, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745602381222, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6447000563144684, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.6447000563144684}}
-:::MLLOG {"namespace": "", "time_ms": 1745602660370, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745602664997, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745602664997, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6944801509380341, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.6944801509380341}}
-:::MLLOG {"namespace": "", "time_ms": 1745602944009, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745602948636, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745602948636, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7051402449607849, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7051402449607849}}
-:::MLLOG {"namespace": "", "time_ms": 1745603227727, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745603232331, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745603232331, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7089028596878052, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7089028596878052}}
-:::MLLOG {"namespace": "", "time_ms": 1745603511699, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745603516332, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745603516332, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7109102308750153, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7109102308750153}}
-:::MLLOG {"namespace": "", "time_ms": 1745603795809, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745603800431, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745603800432, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7122457921504974, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7122457921504974}}
-:::MLLOG {"namespace": "", "time_ms": 1745604080028, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745604084645, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745604084645, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7131429493427277, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7131429493427277}}
-:::MLLOG {"namespace": "", "time_ms": 1745604364612, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745604369261, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745604369261, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7144218623638153, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7144218623638153}}
-:::MLLOG {"namespace": "", "time_ms": 1745604649086, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745604653711, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745604653711, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7148768424987793, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7148768424987793}}
-:::MLLOG {"namespace": "", "time_ms": 1745604935309, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745604939938, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745604939938, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7154791951179504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.7154791951179504}}
-:::MLLOG {"namespace": "", "time_ms": 1745605219927, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745605224572, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745605224572, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7161833882331848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7161833882331848}}
-:::MLLOG {"namespace": "", "time_ms": 1745605510499, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745605515142, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745605515142, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7165493428707123, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7165493428707123}}
-:::MLLOG {"namespace": "", "time_ms": 1745605794599, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745605799240, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745605799241, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7170698583126068, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7170698583126068}}
-:::MLLOG {"namespace": "", "time_ms": 1745606082218, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745606086859, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745606086860, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182413637638092, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7182413637638092}}
-:::MLLOG {"namespace": "", "time_ms": 1745606370703, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2840576, "step_num": 2774}}
-:::MLLOG {"namespace": "", "time_ms": 1745606375349, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2840576, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2840576, "step_num": 2774, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745606375350, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7183951079845429, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2840576, "masked_lm_accuracy": 0.7183951079845429}}
-:::MLLOG {"namespace": "", "time_ms": 1745606655495, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2990080, "step_num": 2920}}
-:::MLLOG {"namespace": "", "time_ms": 1745606660147, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2990080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2990080, "step_num": 2920, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745606660147, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7185700833797455, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2990080, "masked_lm_accuracy": 0.7185700833797455}}
-:::MLLOG {"namespace": "", "time_ms": 1745606941853, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3139584, "step_num": 3066}}
-:::MLLOG {"namespace": "", "time_ms": 1745606946474, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3139584, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3139584, "step_num": 3066, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745606946474, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719360601902008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3139584, "masked_lm_accuracy": 0.719360601902008}}
-:::MLLOG {"namespace": "", "time_ms": 1745607229955, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3289088, "step_num": 3212}}
-:::MLLOG {"namespace": "", "time_ms": 1745607234597, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3289088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3289088, "step_num": 3212, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745607234597, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198850989341736, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3289088, "masked_lm_accuracy": 0.7198850989341736}}
-:::MLLOG {"namespace": "", "time_ms": 1745607515289, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3438592, "step_num": 3358}}
-:::MLLOG {"namespace": "", "time_ms": 1745607519928, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3438592, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3438592, "step_num": 3358, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745607519928, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198968529701233, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3438592, "masked_lm_accuracy": 0.7198968529701233}}
-:::MLLOG {"namespace": "", "time_ms": 1745607800552, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3588096, "step_num": 3504}}
-:::MLLOG {"namespace": "", "time_ms": 1745607805193, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3588096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3588096, "step_num": 3504, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745607805194, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7202662408351899, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3588096, "masked_lm_accuracy": 0.7202662408351899}}
-:::MLLOG {"namespace": "", "time_ms": 1745607805194, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3588096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3588096}}
-:::MLLOG {"namespace": "", "time_ms": 1745607805194, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,84 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745607853432, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853448, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_8xMI300X", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853448, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853448, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853448, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853489, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745607853490, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745609448260, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745609456393, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745609456411, "event_type": "POINT_IN_TIME", "key": "seed", "value": 24956, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468886, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 1024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468886, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468886, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468886, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.0011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.60466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.85437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468887, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 3900, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468888, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468888, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10240, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745609468888, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3993600, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745609497729, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745610146211, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149504, "step_num": 146}}
-:::MLLOG {"namespace": "", "time_ms": 1745610186645, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149504, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149504, "step_num": 146, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745610186646, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37167071998119355, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149504, "masked_lm_accuracy": 0.37167071998119355}}
-:::MLLOG {"namespace": "", "time_ms": 1745610462149, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299008, "step_num": 292}}
-:::MLLOG {"namespace": "", "time_ms": 1745610466711, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299008, "step_num": 292, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745610466711, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3860401749610901, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299008, "masked_lm_accuracy": 0.3860401749610901}}
-:::MLLOG {"namespace": "", "time_ms": 1745610742915, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 448512, "step_num": 438}}
-:::MLLOG {"namespace": "", "time_ms": 1745610747488, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 448512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 448512, "step_num": 438, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745610747488, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5218028664588928, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 448512, "masked_lm_accuracy": 0.5218028664588928}}
-:::MLLOG {"namespace": "", "time_ms": 1745611023698, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 598016, "step_num": 584}}
-:::MLLOG {"namespace": "", "time_ms": 1745611028274, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 598016, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 598016, "step_num": 584, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745611028275, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.673934280872345, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 598016, "masked_lm_accuracy": 0.673934280872345}}
-:::MLLOG {"namespace": "", "time_ms": 1745611304381, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 747520, "step_num": 730}}
-:::MLLOG {"namespace": "", "time_ms": 1745611308983, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 747520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 747520, "step_num": 730, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745611308983, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7041513025760651, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 747520, "masked_lm_accuracy": 0.7041513025760651}}
-:::MLLOG {"namespace": "", "time_ms": 1745611585202, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 897024, "step_num": 876}}
-:::MLLOG {"namespace": "", "time_ms": 1745611589795, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 897024, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 897024, "step_num": 876, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745611589795, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7099382877349854, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 897024, "masked_lm_accuracy": 0.7099382877349854}}
-:::MLLOG {"namespace": "", "time_ms": 1745611866035, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1046528, "step_num": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745611870610, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1046528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1046528, "step_num": 1022, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745611870610, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7122330486774444, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1046528, "masked_lm_accuracy": 0.7122330486774444}}
-:::MLLOG {"namespace": "", "time_ms": 1745612146933, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1196032, "step_num": 1168}}
-:::MLLOG {"namespace": "", "time_ms": 1745612151528, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1196032, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1196032, "step_num": 1168, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745612151529, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7136648654937744, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1196032, "masked_lm_accuracy": 0.7136648654937744}}
-:::MLLOG {"namespace": "", "time_ms": 1745612428104, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1345536, "step_num": 1314}}
-:::MLLOG {"namespace": "", "time_ms": 1745612432695, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1345536, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1345536, "step_num": 1314, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745612432695, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7144280135631561, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1345536, "masked_lm_accuracy": 0.7144280135631561}}
-:::MLLOG {"namespace": "", "time_ms": 1745612709067, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1495040, "step_num": 1460}}
-:::MLLOG {"namespace": "", "time_ms": 1745612713674, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1495040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1495040, "step_num": 1460, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745612713675, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147453427314758, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1495040, "masked_lm_accuracy": 0.7147453427314758}}
-:::MLLOG {"namespace": "", "time_ms": 1745612989926, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1644544, "step_num": 1606}}
-:::MLLOG {"namespace": "", "time_ms": 1745612994505, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1644544, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1644544, "step_num": 1606, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745612994505, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162890911102295, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1644544, "masked_lm_accuracy": 0.7162890911102295}}
-:::MLLOG {"namespace": "", "time_ms": 1745613274460, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1794048, "step_num": 1752}}
-:::MLLOG {"namespace": "", "time_ms": 1745613279032, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1794048, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1794048, "step_num": 1752, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745613279032, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7166337609291077, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1794048, "masked_lm_accuracy": 0.7166337609291077}}
-:::MLLOG {"namespace": "", "time_ms": 1745613554938, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1943552, "step_num": 1898}}
-:::MLLOG {"namespace": "", "time_ms": 1745613559537, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1943552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1943552, "step_num": 1898, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745613559537, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172951698303223, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1943552, "masked_lm_accuracy": 0.7172951698303223}}
-:::MLLOG {"namespace": "", "time_ms": 1745613842482, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2093056, "step_num": 2044}}
-:::MLLOG {"namespace": "", "time_ms": 1745613847063, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2093056, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2093056, "step_num": 2044, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745613847063, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718510490655899, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2093056, "masked_lm_accuracy": 0.718510490655899}}
-:::MLLOG {"namespace": "", "time_ms": 1745614126049, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2242560, "step_num": 2190}}
-:::MLLOG {"namespace": "", "time_ms": 1745614130639, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2242560, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2242560, "step_num": 2190, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745614130639, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7184054613113403, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2242560, "masked_lm_accuracy": 0.7184054613113403}}
-:::MLLOG {"namespace": "", "time_ms": 1745614406167, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2392064, "step_num": 2336}}
-:::MLLOG {"namespace": "", "time_ms": 1745614410769, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2392064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2392064, "step_num": 2336, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745614410769, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189517140388488, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2392064, "masked_lm_accuracy": 0.7189517140388488}}
-:::MLLOG {"namespace": "", "time_ms": 1745614686010, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2541568, "step_num": 2482}}
-:::MLLOG {"namespace": "", "time_ms": 1745614690621, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2541568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2541568, "step_num": 2482, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745614690622, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191794335842132, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2541568, "masked_lm_accuracy": 0.7191794335842132}}
-:::MLLOG {"namespace": "", "time_ms": 1745614968159, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2691072, "step_num": 2628}}
-:::MLLOG {"namespace": "", "time_ms": 1745614972735, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2691072, "step_num": 2628, "samples_count": 10240}}
-:::MLLOG {"namespace": "", "time_ms": 1745614972736, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7202518999576568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2691072, "masked_lm_accuracy": 0.7202518999576568}}
-:::MLLOG {"namespace": "", "time_ms": 1745614972736, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2691072, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2691072}}
-:::MLLOG {"namespace": "", "time_ms": 1745614972736, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,96 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745542295813, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745542295826, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745542295827, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745542295827, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745542295827, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745542296498, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745542296499, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745543895182, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745543905464, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745543905478, "event_type": "POINT_IN_TIME", "key": "seed", "value": 13785, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921082, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921082, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921082, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921083, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745543921084, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745543978728, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745544850167, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745544889032, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745544889032, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37353726426760353, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37353726426760353}}
-:::MLLOG {"namespace": "", "time_ms": 1745545487519, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745545498684, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745545498684, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.41237829072134835, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.41237829072134835}}
-:::MLLOG {"namespace": "", "time_ms": 1745546090548, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745546100134, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745546100135, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5034303157102494, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.5034303157102494}}
-:::MLLOG {"namespace": "", "time_ms": 1745546693196, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745546702828, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745546702828, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6373415260087876, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6373415260087876}}
-:::MLLOG {"namespace": "", "time_ms": 1745547295519, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745547305142, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745547305142, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6955064086686997, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6955064086686997}}
-:::MLLOG {"namespace": "", "time_ms": 1745547896140, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745547905833, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745547905833, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7047913125583104, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7047913125583104}}
-:::MLLOG {"namespace": "", "time_ms": 1745548498633, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745548508143, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745548508143, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7078574929918562, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7078574929918562}}
-:::MLLOG {"namespace": "", "time_ms": 1745549101227, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745549110866, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745549110866, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7101279389290582, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7101279389290582}}
-:::MLLOG {"namespace": "", "time_ms": 1745549701586, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745549711083, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745549711084, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7115791564895994, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7115791564895994}}
-:::MLLOG {"namespace": "", "time_ms": 1745550303748, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745550313288, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745550313288, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.713118288630531, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.713118288630531}}
-:::MLLOG {"namespace": "", "time_ms": 1745550906395, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745550916128, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745550916128, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7131774215471177, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7131774215471177}}
-:::MLLOG {"namespace": "", "time_ms": 1745551506889, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745551516739, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745551516739, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7145353862217494, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7145353862217494}}
-:::MLLOG {"namespace": "", "time_ms": 1745552113873, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745552123576, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745552123576, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7154489999725705, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7154489999725705}}
-:::MLLOG {"namespace": "", "time_ms": 1745552716142, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745552725770, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745552725771, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163924120721363, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7163924120721363}}
-:::MLLOG {"namespace": "", "time_ms": 1745553324307, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745553333814, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745553333815, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716705474399385, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.716705474399385}}
-:::MLLOG {"namespace": "", "time_ms": 1745553926317, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745553936094, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745553936094, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7176795959472656, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7176795959472656}}
-:::MLLOG {"namespace": "", "time_ms": 1745554531890, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745554542845, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745554542846, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178676713080633, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7178676713080633}}
-:::MLLOG {"namespace": "", "time_ms": 1745555133528, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745555143101, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745555143101, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7184804496311006, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7184804496311006}}
-:::MLLOG {"namespace": "", "time_ms": 1745555747973, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745555757481, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745555757481, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718643741948264, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.718643741948264}}
-:::MLLOG {"namespace": "", "time_ms": 1745556357940, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745556367552, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745556367552, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190265468188695, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7190265468188695}}
-:::MLLOG {"namespace": "", "time_ms": 1745556962657, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745556972166, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745556972166, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7194553653399149, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7194553653399149}}
-:::MLLOG {"namespace": "", "time_ms": 1745557564941, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3298944, "step_num": 34364}}
-:::MLLOG {"namespace": "", "time_ms": 1745557574552, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3298944, "step_num": 34364, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745557574552, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.720384947458903, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3298944, "masked_lm_accuracy": 0.720384947458903}}
-:::MLLOG {"namespace": "", "time_ms": 1745557574552, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3298944}}
-:::MLLOG {"namespace": "", "time_ms": 1745557574552, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,96 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745557594256, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594270, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594270, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594270, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594270, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594941, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745557594941, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745559214897, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745559225413, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745559225427, "event_type": "POINT_IN_TIME", "key": "seed", "value": 32312, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240479, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240480, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240481, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240482, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745559240482, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745559294576, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745560166054, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745560205220, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745560205220, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37391537456285384, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37391537456285384}}
-:::MLLOG {"namespace": "", "time_ms": 1745560805893, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745560815463, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745560815463, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4109735344137464, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4109735344137464}}
-:::MLLOG {"namespace": "", "time_ms": 1745561410351, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745561419975, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745561419975, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.49219877805028645, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.49219877805028645}}
-:::MLLOG {"namespace": "", "time_ms": 1745562014334, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745562023781, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745562023781, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6128720873878115, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6128720873878115}}
-:::MLLOG {"namespace": "", "time_ms": 1745562618009, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745562627536, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745562627537, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.697400647117978, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.697400647117978}}
-:::MLLOG {"namespace": "", "time_ms": 1745563220282, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745563229809, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745563229810, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.705579203651065, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.705579203651065}}
-:::MLLOG {"namespace": "", "time_ms": 1745563823608, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745563833062, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745563833062, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7082960753213792, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7082960753213792}}
-:::MLLOG {"namespace": "", "time_ms": 1745564427147, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745564436605, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745564436605, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7107288156236921, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7107288156236921}}
-:::MLLOG {"namespace": "", "time_ms": 1745565028811, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745565038436, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745565038436, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7119181950887045, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7119181950887045}}
-:::MLLOG {"namespace": "", "time_ms": 1745565632265, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745565641826, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745565641827, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7127573898860387, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7127573898860387}}
-:::MLLOG {"namespace": "", "time_ms": 1745566234178, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745566245324, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745566245325, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7137119378362383, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7137119378362383}}
-:::MLLOG {"namespace": "", "time_ms": 1745566837891, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745566847554, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745566847554, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7144972681999207, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7144972681999207}}
-:::MLLOG {"namespace": "", "time_ms": 1745567441551, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745567451141, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745567451141, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152219454447428, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7152219454447428}}
-:::MLLOG {"namespace": "", "time_ms": 1745568051665, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745568061105, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745568061106, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7158037361644564, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7158037361644564}}
-:::MLLOG {"namespace": "", "time_ms": 1745568652783, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745568662536, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745568662536, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160673771585737, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7160673771585737}}
-:::MLLOG {"namespace": "", "time_ms": 1745569255983, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745569265537, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745569265537, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716984482606252, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.716984482606252}}
-:::MLLOG {"namespace": "", "time_ms": 1745569865307, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745569874760, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745569874760, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7171691741262164, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7171691741262164}}
-:::MLLOG {"namespace": "", "time_ms": 1745570473145, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745570482762, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745570482763, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7180932062012809, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7180932062012809}}
-:::MLLOG {"namespace": "", "time_ms": 1745571082302, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745571091949, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745571091950, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7185394008954366, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7185394008954366}}
-:::MLLOG {"namespace": "", "time_ms": 1745571701179, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745571710616, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745571710616, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718915491444724, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.718915491444724}}
-:::MLLOG {"namespace": "", "time_ms": 1745572308022, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745572317478, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745572317479, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7197886319387526, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7197886319387526}}
-:::MLLOG {"namespace": "", "time_ms": 1745572911332, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3298944, "step_num": 34364}}
-:::MLLOG {"namespace": "", "time_ms": 1745572920792, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3298944, "step_num": 34364, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745572920792, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7201146880785624, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3298944, "masked_lm_accuracy": 0.7201146880785624}}
-:::MLLOG {"namespace": "", "time_ms": 1745572920793, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3298944}}
-:::MLLOG {"namespace": "", "time_ms": 1745572920793, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745572941616, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745572941629, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745572941630, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745572941630, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745572941630, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745572942291, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745572942291, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745574879961, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745574890440, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745574890454, "event_type": "POINT_IN_TIME", "key": "seed", "value": 6986, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906135, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906135, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906135, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906135, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906135, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906136, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906137, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906137, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906137, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745574906137, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745574955636, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745575830554, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745575870376, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745575870377, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3737282611074902, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3737282611074902}}
-:::MLLOG {"namespace": "", "time_ms": 1745576471845, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745576481412, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745576481413, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4033156321162269, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4033156321162269}}
-:::MLLOG {"namespace": "", "time_ms": 1745577076023, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745577085523, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745577085523, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4883529495625269, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.4883529495625269}}
-:::MLLOG {"namespace": "", "time_ms": 1745577681373, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745577691036, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745577691036, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6172424344789414, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6172424344789414}}
-:::MLLOG {"namespace": "", "time_ms": 1745578286633, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745578296292, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745578296292, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6984787407375518, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6984787407375518}}
-:::MLLOG {"namespace": "", "time_ms": 1745578891885, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745578901678, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745578901679, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7075154764311654, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7075154764311654}}
-:::MLLOG {"namespace": "", "time_ms": 1745579495449, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745579505011, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745579505011, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7099487849644253, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7099487849644253}}
-:::MLLOG {"namespace": "", "time_ms": 1745580100680, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745580110198, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745580110198, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7115290715580895, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7115290715580895}}
-:::MLLOG {"namespace": "", "time_ms": 1745580703932, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745580713443, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745580713443, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7128627572740828, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7128627572740828}}
-:::MLLOG {"namespace": "", "time_ms": 1745581308511, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745581318215, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745581318215, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135378882998512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7135378882998512}}
-:::MLLOG {"namespace": "", "time_ms": 1745581913204, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745581922724, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745581922724, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7148078146434965, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7148078146434965}}
-:::MLLOG {"namespace": "", "time_ms": 1745582522356, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745582531872, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745582531872, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152938303493318, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7152938303493318}}
-:::MLLOG {"namespace": "", "time_ms": 1745583125261, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745583134766, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745583134767, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7161467688424247, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7161467688424247}}
-:::MLLOG {"namespace": "", "time_ms": 1745583737025, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745583746516, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745583746516, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.71663555928639, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.71663555928639}}
-:::MLLOG {"namespace": "", "time_ms": 1745584340025, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745584350995, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745584350996, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7171961858159019, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7171961858159019}}
-:::MLLOG {"namespace": "", "time_ms": 1745584955665, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745584965146, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745584965146, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7175660905383882, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7175660905383882}}
-:::MLLOG {"namespace": "", "time_ms": 1745585565937, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745585575442, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745585575443, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181211761065892, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7181211761065892}}
-:::MLLOG {"namespace": "", "time_ms": 1745586170179, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745586179685, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745586179685, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7188269053186689, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7188269053186689}}
-:::MLLOG {"namespace": "", "time_ms": 1745586785370, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745586794992, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745586794993, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191171118191311, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7191171118191311}}
-:::MLLOG {"namespace": "", "time_ms": 1745587394486, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745587403988, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745587403988, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7196645231474014, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7196645231474014}}
-:::MLLOG {"namespace": "", "time_ms": 1745587999303, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745588008978, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745588008978, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200902217910403, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7200902217910403}}
-:::MLLOG {"namespace": "", "time_ms": 1745588008978, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3148992}}
-:::MLLOG {"namespace": "", "time_ms": 1745588008979, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,84 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745588028932, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745588028946, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745588028946, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745588028946, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745588028946, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745588029621, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745588029622, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745589719583, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745589729896, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745589729909, "event_type": "POINT_IN_TIME", "key": "seed", "value": 4120, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745143, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745144, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745145, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745146, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745146, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745589745146, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745589790515, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745590662059, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745590701366, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745590701367, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3738187574204944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3738187574204944}}
-:::MLLOG {"namespace": "", "time_ms": 1745591300740, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745591310217, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745591310217, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.407555852049873, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.407555852049873}}
-:::MLLOG {"namespace": "", "time_ms": 1745591904870, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745591914359, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745591914359, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5020121426809402, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.5020121426809402}}
-:::MLLOG {"namespace": "", "time_ms": 1745592506867, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745592517809, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745592517809, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6646601739383879, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6646601739383879}}
-:::MLLOG {"namespace": "", "time_ms": 1745593110284, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745593120055, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745593120055, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.703122741835458, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.703122741835458}}
-:::MLLOG {"namespace": "", "time_ms": 1745593713739, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745593723210, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745593723211, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7083731492360433, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7083731492360433}}
-:::MLLOG {"namespace": "", "time_ms": 1745594317185, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745594326671, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745594326671, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7107717264266241, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7107717264266241}}
-:::MLLOG {"namespace": "", "time_ms": 1745594919068, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745594928715, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745594928715, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7129231032871065, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7129231032871065}}
-:::MLLOG {"namespace": "", "time_ms": 1745595522298, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745595531920, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745595531920, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7140601788248334, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7140601788248334}}
-:::MLLOG {"namespace": "", "time_ms": 1745596125286, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745596134779, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745596134779, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7149945645105271, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7149945645105271}}
-:::MLLOG {"namespace": "", "time_ms": 1745596732267, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745596741919, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745596741920, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7153055503254845, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7153055503254845}}
-:::MLLOG {"namespace": "", "time_ms": 1745597342010, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745597351718, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745597351718, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162052637054807, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7162052637054807}}
-:::MLLOG {"namespace": "", "time_ms": 1745597945037, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745597954497, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745597954497, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7168858408927917, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7168858408927917}}
-:::MLLOG {"namespace": "", "time_ms": 1745598553278, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745598562756, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745598562756, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181580322129386, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7181580322129386}}
-:::MLLOG {"namespace": "", "time_ms": 1745599155685, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745599165357, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745599165357, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7184280049233209, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7184280049233209}}
-:::MLLOG {"namespace": "", "time_ms": 1745599763495, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745599773182, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745599773182, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189048784119743, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7189048784119743}}
-:::MLLOG {"namespace": "", "time_ms": 1745600378094, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745600387579, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745600387579, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719234371752966, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.719234371752966}}
-:::MLLOG {"namespace": "", "time_ms": 1745600994209, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745601003674, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745601003674, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200033778236026, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7200033778236026}}
-:::MLLOG {"namespace": "", "time_ms": 1745601003674, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2699136}}
-:::MLLOG {"namespace": "", "time_ms": 1745601003675, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745601024651, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745601024665, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745601024665, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745601024665, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745601024665, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745601025341, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745601025341, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745602641535, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745602651782, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745602651795, "event_type": "POINT_IN_TIME", "key": "seed", "value": 31501, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667014, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667014, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667014, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667014, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667014, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667015, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745602667016, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745602720684, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745603593793, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745603633838, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745603633838, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3756687879562378, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3756687879562378}}
-:::MLLOG {"namespace": "", "time_ms": 1745604234034, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745604243699, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745604243700, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.41726759388333273, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.41726759388333273}}
-:::MLLOG {"namespace": "", "time_ms": 1745604836741, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745604846207, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745604846207, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6361151451156253, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.6361151451156253}}
-:::MLLOG {"namespace": "", "time_ms": 1745605440504, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745605449980, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745605449981, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7018973787625631, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.7018973787625631}}
-:::MLLOG {"namespace": "", "time_ms": 1745606043898, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745606053400, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745606053400, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7083071765445528, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7083071765445528}}
-:::MLLOG {"namespace": "", "time_ms": 1745606645367, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745606656313, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745606656313, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7105425567854018, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7105425567854018}}
-:::MLLOG {"namespace": "", "time_ms": 1745607248432, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745607258080, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745607258080, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7116558080627805, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7116558080627805}}
-:::MLLOG {"namespace": "", "time_ms": 1745607851555, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745607861222, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745607861223, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7134888819285802, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7134888819285802}}
-:::MLLOG {"namespace": "", "time_ms": 1745608454820, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745608464280, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745608464280, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7140375443867275, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7140375443867275}}
-:::MLLOG {"namespace": "", "time_ms": 1745609055972, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745609065428, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745609065428, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7148039085524422, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7148039085524422}}
-:::MLLOG {"namespace": "", "time_ms": 1745609658543, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745609668034, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745609668035, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7156602161271232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7156602161271232}}
-:::MLLOG {"namespace": "", "time_ms": 1745610261103, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745610270737, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745610270737, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716008236294701, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.716008236294701}}
-:::MLLOG {"namespace": "", "time_ms": 1745610862227, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745610871748, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745610871748, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7165316888264247, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7165316888264247}}
-:::MLLOG {"namespace": "", "time_ms": 1745611475921, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745611485374, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745611485374, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7171947257859367, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7171947257859367}}
-:::MLLOG {"namespace": "", "time_ms": 1745612083788, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745612094912, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745612094912, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181121451514108, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7181121451514108}}
-:::MLLOG {"namespace": "", "time_ms": 1745612686207, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745612695807, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745612695807, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7187516025134495, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7187516025134495}}
-:::MLLOG {"namespace": "", "time_ms": 1745613288639, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745613298492, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745613298493, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186400470279511, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7186400470279511}}
-:::MLLOG {"namespace": "", "time_ms": 1745613892142, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745613901611, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745613901611, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7196106592814128, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7196106592814128}}
-:::MLLOG {"namespace": "", "time_ms": 1745614498093, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745614507551, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745614507551, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719801956699008, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.719801956699008}}
-:::MLLOG {"namespace": "", "time_ms": 1745615109551, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745615119233, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745615119233, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7204833462124779, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7204833462124779}}
-:::MLLOG {"namespace": "", "time_ms": 1745615119234, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745615119234, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,84 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745615138505, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745615138519, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745615138519, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745615138520, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745615138520, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745615139198, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745615139199, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745616751147, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745616761345, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745616761359, "event_type": "POINT_IN_TIME", "key": "seed", "value": 10057, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776690, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776691, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776692, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776693, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776693, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745616776693, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745616820038, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745617693199, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745617733683, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745617733683, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37397653659184776, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37397653659184776}}
-:::MLLOG {"namespace": "", "time_ms": 1745618335201, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745618344856, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745618344856, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3990937738191514, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.3990937738191514}}
-:::MLLOG {"namespace": "", "time_ms": 1745618939487, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745618948957, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745618948957, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.46658501823743187, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.46658501823743187}}
-:::MLLOG {"namespace": "", "time_ms": 1745619545017, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745619554509, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745619554509, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6130311591284615, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6130311591284615}}
-:::MLLOG {"namespace": "", "time_ms": 1745620150097, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745620159777, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745620159777, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.699844073113941, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.699844073113941}}
-:::MLLOG {"namespace": "", "time_ms": 1745620753981, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745620764905, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745620764906, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7078652035622369, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7078652035622369}}
-:::MLLOG {"namespace": "", "time_ms": 1745621358745, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745621368403, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745621368404, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7104341376395452, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7104341376395452}}
-:::MLLOG {"namespace": "", "time_ms": 1745621963465, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745621973218, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745621973218, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7125999070349194, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7125999070349194}}
-:::MLLOG {"namespace": "", "time_ms": 1745622568888, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745622578553, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745622578553, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7138977993102301, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7138977993102301}}
-:::MLLOG {"namespace": "", "time_ms": 1745623172627, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745623182469, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745623182469, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147745728492737, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7147745728492737}}
-:::MLLOG {"namespace": "", "time_ms": 1745623788549, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745623798248, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745623798248, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7157551918710982, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7157551918710982}}
-:::MLLOG {"namespace": "", "time_ms": 1745624393638, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745624403132, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745624403132, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7161134055682591, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7161134055682591}}
-:::MLLOG {"namespace": "", "time_ms": 1745625005439, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745625014940, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745625014940, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7176308717046466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7176308717046466}}
-:::MLLOG {"namespace": "", "time_ms": 1745625609773, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745625619252, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745625619252, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178728024164835, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7178728024164835}}
-:::MLLOG {"namespace": "", "time_ms": 1745626226550, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745626236225, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745626236226, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182088261558897, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7182088261558897}}
-:::MLLOG {"namespace": "", "time_ms": 1745626829609, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745626839220, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745626839221, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189385493596395, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7189385493596395}}
-:::MLLOG {"namespace": "", "time_ms": 1745627434203, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745627443687, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745627443688, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7195644435428438, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7195644435428438}}
-:::MLLOG {"namespace": "", "time_ms": 1745628036848, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745628047895, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745628047895, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7202766804468064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7202766804468064}}
-:::MLLOG {"namespace": "", "time_ms": 1745628047896, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2699136}}
-:::MLLOG {"namespace": "", "time_ms": 1745628047896, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745628067289, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067303, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067303, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067303, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067303, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067966, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745628067966, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745629666209, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745629677049, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745629677063, "event_type": "POINT_IN_TIME", "key": "seed", "value": 12465, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691974, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691974, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691975, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745629691976, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745629740990, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745630621098, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745630660651, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745630660651, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37397750247092476, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37397750247092476}}
-:::MLLOG {"namespace": "", "time_ms": 1745631266361, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745631276039, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745631276039, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4039459753604162, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4039459753604162}}
-:::MLLOG {"namespace": "", "time_ms": 1745631876011, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745631885494, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745631885494, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.46332124216215953, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.46332124216215953}}
-:::MLLOG {"namespace": "", "time_ms": 1745632483680, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745632493164, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745632493164, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5909533494994754, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.5909533494994754}}
-:::MLLOG {"namespace": "", "time_ms": 1745633092727, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745633102224, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745633102224, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.695507520153409, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.695507520153409}}
-:::MLLOG {"namespace": "", "time_ms": 1745633701690, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745633711166, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745633711166, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.70634758018312, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.70634758018312}}
-:::MLLOG {"namespace": "", "time_ms": 1745634308886, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745634318381, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745634318382, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7093000508490063, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7093000508490063}}
-:::MLLOG {"namespace": "", "time_ms": 1745634917081, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745634926554, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745634926554, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7113453558513096, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7113453558513096}}
-:::MLLOG {"namespace": "", "time_ms": 1745635523693, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745635534738, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745635534738, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7130639910697937, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7130639910697937}}
-:::MLLOG {"namespace": "", "time_ms": 1745636131809, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745636141329, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745636141330, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7139686294964381, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7139686294964381}}
-:::MLLOG {"namespace": "", "time_ms": 1745636739800, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745636749545, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745636749545, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7148919718606132, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7148919718606132}}
-:::MLLOG {"namespace": "", "time_ms": 1745637359823, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745637369305, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745637369306, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7157319557099115, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7157319557099115}}
-:::MLLOG {"namespace": "", "time_ms": 1745637971557, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745637981047, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745637981047, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163103591828119, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7163103591828119}}
-:::MLLOG {"namespace": "", "time_ms": 1745638579005, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745638588492, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745638588493, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172297625314622, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7172297625314622}}
-:::MLLOG {"namespace": "", "time_ms": 1745639192542, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745639202337, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745639202338, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7177095975194658, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7177095975194658}}
-:::MLLOG {"namespace": "", "time_ms": 1745639798352, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745639807842, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745639807842, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181210790361677, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7181210790361677}}
-:::MLLOG {"namespace": "", "time_ms": 1745640406137, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745640415674, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745640415674, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7187060095015027, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7187060095015027}}
-:::MLLOG {"namespace": "", "time_ms": 1745641012089, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745641021561, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745641021562, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191783354395912, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7191783354395912}}
-:::MLLOG {"namespace": "", "time_ms": 1745641631830, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745641641317, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745641641317, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7195633859861464, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7195633859861464}}
-:::MLLOG {"namespace": "", "time_ms": 1745642245529, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745642255025, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745642255026, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7203016593342736, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7203016593342736}}
-:::MLLOG {"namespace": "", "time_ms": 1745642255026, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745642255026, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,102 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745642274895, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745642274909, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745642274909, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745642274909, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745642274909, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745642275592, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745642275592, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745643905216, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745643915672, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745643915685, "event_type": "POINT_IN_TIME", "key": "seed", "value": 11730, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930970, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930970, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930970, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930971, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745643930972, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745643991209, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745644864276, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745644905184, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745644905184, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3781838553292411, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3781838553292411}}
-:::MLLOG {"namespace": "", "time_ms": 1745645504653, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745645514115, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745645514116, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4123739058063144, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4123739058063144}}
-:::MLLOG {"namespace": "", "time_ms": 1745646105727, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745646115516, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745646115516, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4941163951442355, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.4941163951442355}}
-:::MLLOG {"namespace": "", "time_ms": 1745646708029, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745646717701, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745646717701, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5998574114981152, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.5998574114981152}}
-:::MLLOG {"namespace": "", "time_ms": 1745647310461, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745647319938, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745647319938, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6882066153344654, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6882066153344654}}
-:::MLLOG {"namespace": "", "time_ms": 1745647910916, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745647920440, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745647920441, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7018242018563406, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7018242018563406}}
-:::MLLOG {"namespace": "", "time_ms": 1745648513412, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745648522931, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745648522932, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7060379794665745, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7060379794665745}}
-:::MLLOG {"namespace": "", "time_ms": 1745649115751, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745649125216, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745649125217, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7098249503544398, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7098249503544398}}
-:::MLLOG {"namespace": "", "time_ms": 1745649715710, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745649725226, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745649725226, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.710793800013406, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.710793800013406}}
-:::MLLOG {"namespace": "", "time_ms": 1745650317598, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745650327086, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745650327086, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7119041085243225, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7119041085243225}}
-:::MLLOG {"namespace": "", "time_ms": 1745650919471, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745650928949, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745650928949, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7127983939080011, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7127983939080011}}
-:::MLLOG {"namespace": "", "time_ms": 1745651519561, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745651529185, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745651529186, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7138419349988302, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7138419349988302}}
-:::MLLOG {"namespace": "", "time_ms": 1745652122200, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745652131782, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745652131782, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147189242499216, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7147189242499216}}
-:::MLLOG {"namespace": "", "time_ms": 1745652730615, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745652740312, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745652740312, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152118898573376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7152118898573376}}
-:::MLLOG {"namespace": "", "time_ms": 1745653331444, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745653340998, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745653340998, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.715751085962568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.715751085962568}}
-:::MLLOG {"namespace": "", "time_ms": 1745653933389, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745653942898, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745653942898, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.71591789268312, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.71591789268312}}
-:::MLLOG {"namespace": "", "time_ms": 1745654540794, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745654550615, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745654550616, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716970343816848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.716970343816848}}
-:::MLLOG {"namespace": "", "time_ms": 1745655142498, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745655151972, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745655151972, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7177020226206098, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7177020226206098}}
-:::MLLOG {"namespace": "", "time_ms": 1745655742995, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745655752627, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745655752627, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181210052399408, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7181210052399408}}
-:::MLLOG {"namespace": "", "time_ms": 1745656351799, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745656361268, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745656361268, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181035654885428, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7181035654885428}}
-:::MLLOG {"namespace": "", "time_ms": 1745656967639, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745656978577, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745656978577, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189842661221822, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7189842661221822}}
-:::MLLOG {"namespace": "", "time_ms": 1745657575734, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3298944, "step_num": 34364}}
-:::MLLOG {"namespace": "", "time_ms": 1745657585375, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3298944, "step_num": 34364, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745657585375, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7194626552718026, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3298944, "masked_lm_accuracy": 0.7194626552718026}}
-:::MLLOG {"namespace": "", "time_ms": 1745658178120, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3448896, "step_num": 35926}}
-:::MLLOG {"namespace": "", "time_ms": 1745658187683, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3448896, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3448896, "step_num": 35926, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745658187683, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7197951980999537, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3448896, "masked_lm_accuracy": 0.7197951980999537}}
-:::MLLOG {"namespace": "", "time_ms": 1745658780502, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3598848, "step_num": 37488}}
-:::MLLOG {"namespace": "", "time_ms": 1745658789976, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3598848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3598848, "step_num": 37488, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745658789976, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200545878637404, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3598848, "masked_lm_accuracy": 0.7200545878637404}}
-:::MLLOG {"namespace": "", "time_ms": 1745658789977, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3598848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3598848}}
-:::MLLOG {"namespace": "", "time_ms": 1745658789977, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,84 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745658809591, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745658809604, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745658809604, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745658809604, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745658809604, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745658810272, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745658810273, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745660394086, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745660404397, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745660404410, "event_type": "POINT_IN_TIME", "key": "seed", "value": 30006, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419324, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419324, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419324, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419324, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419325, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745660419326, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745660468452, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745661348515, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745661388786, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745661388787, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37456812262535094, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37456812262535094}}
-:::MLLOG {"namespace": "", "time_ms": 1745661996315, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745662005821, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745662005821, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4560752814724332, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4560752814724332}}
-:::MLLOG {"namespace": "", "time_ms": 1745662609218, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745662618735, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745662618736, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6749432524045308, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.6749432524045308}}
-:::MLLOG {"namespace": "", "time_ms": 1745663219268, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745663228798, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745663228798, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7049068002473741, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.7049068002473741}}
-:::MLLOG {"namespace": "", "time_ms": 1745663830866, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745663840454, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745663840455, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7094300275757199, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7094300275757199}}
-:::MLLOG {"namespace": "", "time_ms": 1745664440883, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745664450546, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745664450546, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7118306568690709, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7118306568690709}}
-:::MLLOG {"namespace": "", "time_ms": 1745665052620, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745665062331, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745665062331, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7131405120804196, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7131405120804196}}
-:::MLLOG {"namespace": "", "time_ms": 1745665664579, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745665674079, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745665674079, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7141566656884692, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7141566656884692}}
-:::MLLOG {"namespace": "", "time_ms": 1745666274721, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745666285824, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745666285824, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.714853200458345, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.714853200458345}}
-:::MLLOG {"namespace": "", "time_ms": 1745666886316, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745666895839, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745666895839, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7154801947729929, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7154801947729929}}
-:::MLLOG {"namespace": "", "time_ms": 1745667505514, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745667515034, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745667515034, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160271604855856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7160271604855856}}
-:::MLLOG {"namespace": "", "time_ms": 1745668117399, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745668126911, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745668126912, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7161095375106448, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7161095375106448}}
-:::MLLOG {"namespace": "", "time_ms": 1745668727427, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745668736920, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745668736921, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172707217080253, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7172707217080253}}
-:::MLLOG {"namespace": "", "time_ms": 1745669347896, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745669357386, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745669357386, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181751222837539, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7181751222837539}}
-:::MLLOG {"namespace": "", "time_ms": 1745669959126, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745669968787, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745669968787, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718704723176502, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.718704723176502}}
-:::MLLOG {"namespace": "", "time_ms": 1745670573606, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745670583206, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745670583206, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7194259950092861, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7194259950092861}}
-:::MLLOG {"namespace": "", "time_ms": 1745671185025, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745671194623, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745671194623, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.71981109891619, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.71981109891619}}
-:::MLLOG {"namespace": "", "time_ms": 1745671799743, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745671810731, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745671810731, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7204012291772025, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7204012291772025}}
-:::MLLOG {"namespace": "", "time_ms": 1745671810731, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2699136}}
-:::MLLOG {"namespace": "", "time_ms": 1745671810732, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745671829193, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829206, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829207, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829207, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829207, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829888, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745671829889, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745673412514, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745673423084, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745673423098, "event_type": "POINT_IN_TIME", "key": "seed", "value": 27130, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438047, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438047, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438047, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438047, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438048, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745673438049, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745673490539, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745674363852, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745674404525, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745674404526, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37596466257458644, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37596466257458644}}
-:::MLLOG {"namespace": "", "time_ms": 1745675008663, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745675018130, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745675018130, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4192629811309633, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4192629811309633}}
-:::MLLOG {"namespace": "", "time_ms": 1745675616281, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745675625972, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745675625973, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5021054284913199, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.5021054284913199}}
-:::MLLOG {"namespace": "", "time_ms": 1745676222268, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745676232040, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745676232041, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6380213788577489, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6380213788577489}}
-:::MLLOG {"namespace": "", "time_ms": 1745676829655, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745676839323, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745676839324, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7022705288160415, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7022705288160415}}
-:::MLLOG {"namespace": "", "time_ms": 1745677435100, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745677446056, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745677446057, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7073904724348159, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7073904724348159}}
-:::MLLOG {"namespace": "", "time_ms": 1745678041778, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745678051320, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745678051320, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7095591738110497, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7095591738110497}}
-:::MLLOG {"namespace": "", "time_ms": 1745678648680, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745678658179, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745678658180, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7115420869418553, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7115420869418553}}
-:::MLLOG {"namespace": "", "time_ms": 1745679256309, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745679265943, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745679265943, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7130387618428184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7130387618428184}}
-:::MLLOG {"namespace": "", "time_ms": 1745679861819, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745679871292, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745679871292, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135955617541359, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7135955617541359}}
-:::MLLOG {"namespace": "", "time_ms": 1745680468796, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745680478297, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745680478297, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7146279448554629, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7146279448554629}}
-:::MLLOG {"namespace": "", "time_ms": 1745681075639, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745681085107, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745681085107, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7151510505449205, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7151510505449205}}
-:::MLLOG {"namespace": "", "time_ms": 1745681680758, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745681690230, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745681690230, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7165615944635301, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7165615944635301}}
-:::MLLOG {"namespace": "", "time_ms": 1745682292031, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745682301481, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745682301482, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7166609871955145, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7166609871955145}}
-:::MLLOG {"namespace": "", "time_ms": 1745682896874, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745682907952, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745682907952, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717455704439254, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.717455704439254}}
-:::MLLOG {"namespace": "", "time_ms": 1745683523972, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745683533688, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745683533689, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7179811846642267, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7179811846642267}}
-:::MLLOG {"namespace": "", "time_ms": 1745684130394, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745684139891, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745684139891, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7185609034129552, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7185609034129552}}
-:::MLLOG {"namespace": "", "time_ms": 1745684736989, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745684746439, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745684746439, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7191170454025269, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7191170454025269}}
-:::MLLOG {"namespace": "", "time_ms": 1745685345638, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745685355274, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745685355275, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.71960460628782, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.71960460628782}}
-:::MLLOG {"namespace": "", "time_ms": 1745685966333, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745685976054, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745685976054, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200535161154611, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7200535161154611}}
-:::MLLOG {"namespace": "", "time_ms": 1745685976055, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745685976055, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,46 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745596628137, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 382}}
-:::MLLOG {"namespace": "", "time_ms": 1745596628178, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 383}}
-:::MLLOG {"namespace": "", "time_ms": 1745596628178, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 384}}
-:::MLLOG {"namespace": "", "time_ms": 1745596628178, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 385}}
-:::MLLOG {"namespace": "", "time_ms": 1745596628178, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "retinanet", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 387}}
-:::MLLOG {"namespace": "", "time_ms": 1745596629954, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 390}}
-:::MLLOG {"namespace": "", "time_ms": 1745596629955, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 391}}
-:::MLLOG {"namespace": "", "time_ms": 1745598065772, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 658}}
-:::MLLOG {"namespace": "", "time_ms": 1745598081470, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 394}}
-:::MLLOG {"namespace": "", "time_ms": 1745598081512, "event_type": "POINT_IN_TIME", "key": "seed", "value": 23282, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 395}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088273, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 504}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 12191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 505}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 259, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 506}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "epoch_count", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 507}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "first_epoch_num", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 508}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "adam", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 510}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 9.5e-05, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 511}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "opt_weight_decay", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 512}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088274, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_epochs", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 513}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088275, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_factor", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 514}}
-:::MLLOG {"namespace": "", "time_ms": 1745598088275, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 515}}
-:::MLLOG {"namespace": "", "time_ms": 1745598144406, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 527, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745605078062, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 598, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745605078063, "event_type": "INTERVAL_START", "key": "eval_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 603, "epoch_num": 1}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745610378469, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.2608930553164607, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 679, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745610378469, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 680, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745610378469, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 527, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745616941326, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 598, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745616941327, "event_type": "INTERVAL_START", "key": "eval_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 603, "epoch_num": 2}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745622185857, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.31207695716564665, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 679, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745622185858, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 680, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745622185858, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 527, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745628713800, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 598, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745628713800, "event_type": "INTERVAL_START", "key": "eval_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 603, "epoch_num": 3}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745633828548, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.32695300496649193, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 679, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745633828548, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 680, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745633828549, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 527, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745640403678, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 598, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745640403679, "event_type": "INTERVAL_START", "key": "eval_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 603, "epoch_num": 4}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745645485614, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.34190927146960864, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 679, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745645485615, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 680, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745645485615, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 686, "status": "success"}}
@@ -1,46 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745708052929, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 380}}
-:::MLLOG {"namespace": "", "time_ms": 1745708052970, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 381}}
-:::MLLOG {"namespace": "", "time_ms": 1745708052970, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 382}}
-:::MLLOG {"namespace": "", "time_ms": 1745708052970, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 383}}
-:::MLLOG {"namespace": "", "time_ms": 1745708052970, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "retinanet", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 385}}
-:::MLLOG {"namespace": "", "time_ms": 1745708055312, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 388}}
-:::MLLOG {"namespace": "", "time_ms": 1745708055312, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 389}}
-:::MLLOG {"namespace": "", "time_ms": 1745709484510, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 656}}
-:::MLLOG {"namespace": "", "time_ms": 1745709499880, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 392}}
-:::MLLOG {"namespace": "", "time_ms": 1745709499922, "event_type": "POINT_IN_TIME", "key": "seed", "value": 3218, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 393}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506804, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 502}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506805, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 12191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 503}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506805, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 259, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 504}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506805, "event_type": "POINT_IN_TIME", "key": "epoch_count", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 505}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "first_epoch_num", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 506}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "adam", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 508}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 9.5e-05, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 509}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "opt_weight_decay", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 510}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_epochs", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 511}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_factor", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 512}}
-:::MLLOG {"namespace": "", "time_ms": 1745709506806, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 513}}
-:::MLLOG {"namespace": "", "time_ms": 1745709564057, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745716423332, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745716423333, "event_type": "INTERVAL_START", "key": "eval_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 1}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745721892086, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.2644758301871188, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745721892087, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745721892087, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745728717917, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745728717918, "event_type": "INTERVAL_START", "key": "eval_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 2}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745734129092, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3183940553292647, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745734129092, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745734129092, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745740758848, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745740758849, "event_type": "INTERVAL_START", "key": "eval_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 3}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745746017219, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.330829179299047, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745746017219, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745746017219, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745752685505, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745752685506, "event_type": "INTERVAL_START", "key": "eval_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 4}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745757915230, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3430538198992862, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745757915231, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745757915231, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 684, "status": "success"}}
@@ -1,46 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745757942370, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 380}}
-:::MLLOG {"namespace": "", "time_ms": 1745757942411, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 381}}
-:::MLLOG {"namespace": "", "time_ms": 1745757942411, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 382}}
-:::MLLOG {"namespace": "", "time_ms": 1745757942411, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 383}}
-:::MLLOG {"namespace": "", "time_ms": 1745757942411, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "retinanet", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 385}}
-:::MLLOG {"namespace": "", "time_ms": 1745757943058, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 388}}
-:::MLLOG {"namespace": "", "time_ms": 1745757943059, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 389}}
-:::MLLOG {"namespace": "", "time_ms": 1745759379793, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 656}}
-:::MLLOG {"namespace": "", "time_ms": 1745759394363, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 392}}
-:::MLLOG {"namespace": "", "time_ms": 1745759394404, "event_type": "POINT_IN_TIME", "key": "seed", "value": 7068, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 393}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401265, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 502}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401265, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 12191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 503}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 259, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 504}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "epoch_count", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 505}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "first_epoch_num", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 506}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "adam", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 508}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 9.5e-05, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 509}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "opt_weight_decay", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 510}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401266, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_epochs", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 511}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401267, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_factor", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 512}}
-:::MLLOG {"namespace": "", "time_ms": 1745759401267, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 513}}
-:::MLLOG {"namespace": "", "time_ms": 1745759458864, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745766229351, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745766229352, "event_type": "INTERVAL_START", "key": "eval_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 1}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745771664180, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.2618442233208197, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745771664180, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745771664181, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745778271730, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745778271731, "event_type": "INTERVAL_START", "key": "eval_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 2}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745783649281, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3050222595524408, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745783649281, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745783649282, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745790269694, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745790269695, "event_type": "INTERVAL_START", "key": "eval_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 3}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745795622203, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.33438554461867026, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745795622204, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745795622204, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745802201592, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745802201593, "event_type": "INTERVAL_START", "key": "eval_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 4}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745807409644, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3413173788267323, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745807409644, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745807409644, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 684, "status": "success"}}
@@ -1,46 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745708740629, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 380}}
-:::MLLOG {"namespace": "", "time_ms": 1745708740670, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 381}}
-:::MLLOG {"namespace": "", "time_ms": 1745708740670, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 382}}
-:::MLLOG {"namespace": "", "time_ms": 1745708740670, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 383}}
-:::MLLOG {"namespace": "", "time_ms": 1745708740670, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "retinanet", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 385}}
-:::MLLOG {"namespace": "", "time_ms": 1745708753515, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 388}}
-:::MLLOG {"namespace": "", "time_ms": 1745708753515, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 389}}
-:::MLLOG {"namespace": "", "time_ms": 1745710196875, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 656}}
-:::MLLOG {"namespace": "", "time_ms": 1745710211866, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 392}}
-:::MLLOG {"namespace": "", "time_ms": 1745710211906, "event_type": "POINT_IN_TIME", "key": "seed", "value": 1934, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 393}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219928, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 502}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219929, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 12191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 503}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219929, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 259, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 504}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219929, "event_type": "POINT_IN_TIME", "key": "epoch_count", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 505}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219929, "event_type": "POINT_IN_TIME", "key": "first_epoch_num", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 506}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "adam", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 508}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 9.5e-05, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 509}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "opt_weight_decay", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 510}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_epochs", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 511}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_factor", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 512}}
-:::MLLOG {"namespace": "", "time_ms": 1745710219930, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 513}}
-:::MLLOG {"namespace": "", "time_ms": 1745710276595, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745717038732, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745717038733, "event_type": "INTERVAL_START", "key": "eval_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 1}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745722476155, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.24994336549495808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745722476156, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745722476156, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745729177485, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745729177486, "event_type": "INTERVAL_START", "key": "eval_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 2}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745734589630, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.30947442932060776, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745734589630, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745734589630, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745741107714, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745741107714, "event_type": "INTERVAL_START", "key": "eval_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 3}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745746523920, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3304143886715271, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745746523920, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745746523920, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745753134001, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745753134002, "event_type": "INTERVAL_START", "key": "eval_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 4}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745758428287, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3429861420134466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745758428288, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745758428288, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 684, "status": "success"}}
@@ -1,45 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745758455763, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 380}}
-:::MLLOG {"namespace": "", "time_ms": 1745758455804, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_green", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 381}}
-:::MLLOG {"namespace": "", "time_ms": 1745758455804, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 382}}
-:::MLLOG {"namespace": "", "time_ms": 1745758455804, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 383}}
-:::MLLOG {"namespace": "", "time_ms": 1745758455805, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "retinanet", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 385}}
-:::MLLOG {"namespace": "", "time_ms": 1745758457940, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 388}}
-:::MLLOG {"namespace": "", "time_ms": 1745758457941, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 389}}
-:::MLLOG {"namespace": "", "time_ms": 1745759900517, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 656}}
-:::MLLOG {"namespace": "", "time_ms": 1745759915495, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 392}}
-:::MLLOG {"namespace": "", "time_ms": 1745759915536, "event_type": "POINT_IN_TIME", "key": "seed", "value": 25159, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 393}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922365, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 502}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 12191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 503}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 259, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 504}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "epoch_count", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 505}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "first_epoch_num", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 506}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "adam", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 508}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922366, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 9.5e-05, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 509}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922367, "event_type": "POINT_IN_TIME", "key": "opt_weight_decay", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 510}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922367, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_epochs", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 511}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922367, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_factor", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 512}}
-:::MLLOG {"namespace": "", "time_ms": 1745759922367, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 513}}
-:::MLLOG {"namespace": "", "time_ms": 1745759981024, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745766937876, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745766937877, "event_type": "INTERVAL_START", "key": "eval_start", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 1}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745772433927, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.25660616888772175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745772433927, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 1}}
-:::MLLOG {"namespace": "", "time_ms": 1745772433927, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745779249804, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745779249805, "event_type": "INTERVAL_START", "key": "eval_start", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 2}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745784709047, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3114751446994825, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745784709048, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 2}}
-:::MLLOG {"namespace": "", "time_ms": 1745784709048, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745791366481, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745791366482, "event_type": "INTERVAL_START", "key": "eval_start", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 3}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745796796512, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.33395135022162803, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745796796512, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 3}}
-:::MLLOG {"namespace": "", "time_ms": 1745796796512, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 525, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745803562272, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 596, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745803562273, "event_type": "INTERVAL_START", "key": "eval_start", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 601, "epoch_num": 4}}
-
-:::MLLOG {"namespace": "", "time_ms": 1745808971898, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3397162205764848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 677, "epoch_num": 4}}
-:::MLLOG {"namespace": "", "time_ms": 1745808971899, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 4, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 678, "epoch_num": 4}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745787472660, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472675, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472675, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472675, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472675, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472820, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745787472820, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745789021684, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745789063002, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745789063016, "event_type": "POINT_IN_TIME", "key": "seed", "value": 11341, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084054, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084054, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084054, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084054, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084055, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745789084056, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745789130103, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745790243120, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745790285403, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745790285404, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3725697540101551, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3725697540101551}}
-:::MLLOG {"namespace": "", "time_ms": 1745791130486, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745791146651, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745791146651, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4053410061768123, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4053410061768123}}
-:::MLLOG {"namespace": "", "time_ms": 1745791993483, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745792009647, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745792009647, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4546036266145252, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.4546036266145252}}
-:::MLLOG {"namespace": "", "time_ms": 1745792857040, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745792873247, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745792873248, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5833114096096583, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.5833114096096583}}
-:::MLLOG {"namespace": "", "time_ms": 1745793720402, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745793736590, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745793736590, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6891417679332551, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6891417679332551}}
-:::MLLOG {"namespace": "", "time_ms": 1745794586055, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745794602267, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745794602267, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7027380545934041, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7027380545934041}}
-:::MLLOG {"namespace": "", "time_ms": 1745795451897, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745795468074, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745795468074, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7082311675662086, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7082311675662086}}
-:::MLLOG {"namespace": "", "time_ms": 1745796316509, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745796332716, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745796332716, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7108195407049996, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7108195407049996}}
-:::MLLOG {"namespace": "", "time_ms": 1745797182147, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745797198362, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745797198363, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7119770498502822, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7119770498502822}}
-:::MLLOG {"namespace": "", "time_ms": 1745798047629, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745798063852, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745798063852, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7129385369164604, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7129385369164604}}
-:::MLLOG {"namespace": "", "time_ms": 1745798917116, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745798933320, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745798933320, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.713829753512428, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.713829753512428}}
-:::MLLOG {"namespace": "", "time_ms": 1745799781286, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745799797515, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745799797515, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7143420588402521, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7143420588402521}}
-:::MLLOG {"namespace": "", "time_ms": 1745800650547, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745800666777, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745800666777, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7151911968276614, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7151911968276614}}
-:::MLLOG {"namespace": "", "time_ms": 1745801513018, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745801529219, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745801529219, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162183267729623, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7162183267729623}}
-:::MLLOG {"namespace": "", "time_ms": 1745802383777, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745802399993, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745802399993, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163212441262745, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7163212441262745}}
-:::MLLOG {"namespace": "", "time_ms": 1745803252264, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745803268480, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745803268480, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7175917211033049, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7175917211033049}}
-:::MLLOG {"namespace": "", "time_ms": 1745804120412, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745804136630, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745804136630, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178130229314168, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7178130229314168}}
-:::MLLOG {"namespace": "", "time_ms": 1745804990109, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745805006323, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745805006324, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182639241218567, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7182639241218567}}
-:::MLLOG {"namespace": "", "time_ms": 1745805853485, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745805869680, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745805869680, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7185757029624212, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7185757029624212}}
-:::MLLOG {"namespace": "", "time_ms": 1745806722533, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745806738689, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745806738690, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192848898115612, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7192848898115612}}
-:::MLLOG {"namespace": "", "time_ms": 1745807591120, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745807607349, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745807607349, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200041702815465, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7200041702815465}}
-:::MLLOG {"namespace": "", "time_ms": 1745807607350, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3148992}}
-:::MLLOG {"namespace": "", "time_ms": 1745807607350, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745807635676, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635691, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635691, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635691, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635691, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635800, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745807635800, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745809130305, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745809170647, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745809170661, "event_type": "POINT_IN_TIME", "key": "seed", "value": 9893, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187044, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187045, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187045, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187045, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187045, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187045, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187046, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187047, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187047, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187047, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745809187047, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745809235620, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745810359119, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745810401248, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745810401248, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3756656459399632, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3756656459399632}}
-:::MLLOG {"namespace": "", "time_ms": 1745811253490, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745811270105, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745811270105, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.408587916692098, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.408587916692098}}
-:::MLLOG {"namespace": "", "time_ms": 1745812123931, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745812140534, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745812140535, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.48328256692205157, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.48328256692205157}}
-:::MLLOG {"namespace": "", "time_ms": 1745812994623, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745813011247, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745813011247, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6104218710036505, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6104218710036505}}
-:::MLLOG {"namespace": "", "time_ms": 1745813866035, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745813882688, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745813882689, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6986918545904613, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6986918545904613}}
-:::MLLOG {"namespace": "", "time_ms": 1745814739714, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745814756340, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745814756340, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7074495644796462, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7074495644796462}}
-:::MLLOG {"namespace": "", "time_ms": 1745815613451, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745815630089, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745815630090, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7102503475688753, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7102503475688753}}
-:::MLLOG {"namespace": "", "time_ms": 1745816485859, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745816502544, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745816502545, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7119106928507487, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7119106928507487}}
-:::MLLOG {"namespace": "", "time_ms": 1745817359005, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745817375661, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745817375662, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7136906277565729, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7136906277565729}}
-:::MLLOG {"namespace": "", "time_ms": 1745818232065, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745818248716, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745818248716, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7140245165143694, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7140245165143694}}
-:::MLLOG {"namespace": "", "time_ms": 1745819109151, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745819125778, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745819125778, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7150250207810175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7150250207810175}}
-:::MLLOG {"namespace": "", "time_ms": 1745819981082, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745819997739, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745819997739, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7156877937771026, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7156877937771026}}
-:::MLLOG {"namespace": "", "time_ms": 1745820853355, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745820870012, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745820870012, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163892950330462, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7163892950330462}}
-:::MLLOG {"namespace": "", "time_ms": 1745821728724, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745821745361, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745821745361, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172291585377284, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7172291585377284}}
-:::MLLOG {"namespace": "", "time_ms": 1745822604259, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745822620919, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745822620919, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717616525718144, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.717616525718144}}
-:::MLLOG {"namespace": "", "time_ms": 1745823480930, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745823497590, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745823497590, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718469910962241, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.718469910962241}}
-:::MLLOG {"namespace": "", "time_ms": 1745824350574, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745824368623, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745824368624, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.718998844850631, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.718998844850631}}
-:::MLLOG {"namespace": "", "time_ms": 1745825222149, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745825238786, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745825238786, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7196502293859209, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7196502293859209}}
-:::MLLOG {"namespace": "", "time_ms": 1745826105033, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745826121638, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745826121638, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198895817711239, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7198895817711239}}
-:::MLLOG {"namespace": "", "time_ms": 1745826976460, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745826993116, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745826993116, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7203692260242643, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7203692260242643}}
-:::MLLOG {"namespace": "", "time_ms": 1745826993117, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745826993117, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745827020398, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020413, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020413, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020413, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020413, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020525, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745827020525, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745828557523, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745828597922, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745828597937, "event_type": "POINT_IN_TIME", "key": "seed", "value": 11581, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614084, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614084, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614084, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614085, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745828614086, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745828663687, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745829766682, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745829807504, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745829807505, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37525955580529713, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37525955580529713}}
-:::MLLOG {"namespace": "", "time_ms": 1745830642144, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745830658143, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745830658143, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4122683204355694, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4122683204355694}}
-:::MLLOG {"namespace": "", "time_ms": 1745831494398, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745831510372, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745831510372, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.49358772947674706, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.49358772947674706}}
-:::MLLOG {"namespace": "", "time_ms": 1745832347052, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745832363071, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745832363071, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6273789207140604, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6273789207140604}}
-:::MLLOG {"namespace": "", "time_ms": 1745833200443, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745833217885, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745833217885, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7020660332271031, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7020660332271031}}
-:::MLLOG {"namespace": "", "time_ms": 1745834055605, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745834071615, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745834071615, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7076224071638925, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7076224071638925}}
-:::MLLOG {"namespace": "", "time_ms": 1745834910861, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745834926878, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745834926878, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7098002570016043, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7098002570016043}}
-:::MLLOG {"namespace": "", "time_ms": 1745835766033, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745835782039, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745835782039, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7117016775267465, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7117016775267465}}
-:::MLLOG {"namespace": "", "time_ms": 1745836619573, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745836635604, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745836635604, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135275136856806, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7135275136856806}}
-:::MLLOG {"namespace": "", "time_ms": 1745837474097, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745837490110, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745837490110, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7144534780865623, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7144534780865623}}
-:::MLLOG {"namespace": "", "time_ms": 1745838326894, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745838344267, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745838344267, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7147606486365908, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7147606486365908}}
-:::MLLOG {"namespace": "", "time_ms": 1745839180844, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745839196883, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745839196883, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152229150136312, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7152229150136312}}
-:::MLLOG {"namespace": "", "time_ms": 1745840040064, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745840056101, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745840056102, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7165876416932969, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7165876416932969}}
-:::MLLOG {"namespace": "", "time_ms": 1745840891976, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745840909372, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745840909372, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7169663355464027, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7169663355464027}}
-:::MLLOG {"namespace": "", "time_ms": 1745841750085, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745841766074, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745841766074, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178037592342922, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7178037592342922}}
-:::MLLOG {"namespace": "", "time_ms": 1745842606890, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745842622892, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745842622892, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7182614950906663, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7182614950906663}}
-:::MLLOG {"namespace": "", "time_ms": 1745843459937, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745843475947, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745843475948, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189727822939554, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7189727822939554}}
-:::MLLOG {"namespace": "", "time_ms": 1745844322719, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745844338697, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745844338698, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7194424441882542, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7194424441882542}}
-:::MLLOG {"namespace": "", "time_ms": 1745845183319, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745845199341, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745845199342, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7199910112789699, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7199910112789699}}
-:::MLLOG {"namespace": "", "time_ms": 1745846049798, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745846065808, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745846065809, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200804040545509, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7200804040545509}}
-:::MLLOG {"namespace": "", "time_ms": 1745846065809, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745846065809, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745766145092, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145106, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145106, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145106, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145107, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145229, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745766145229, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745767375299, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745767395079, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745767395094, "event_type": "POINT_IN_TIME", "key": "seed", "value": 27065, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411312, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411312, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411312, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411312, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411313, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745767411314, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745767458799, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745768572452, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745768621526, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745768621527, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.373991007180441, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.373991007180441}}
-:::MLLOG {"namespace": "", "time_ms": 1745769469811, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745769486194, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745769486194, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4115104800178891, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4115104800178891}}
-:::MLLOG {"namespace": "", "time_ms": 1745770335677, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745770352069, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745770352069, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5021448158082508, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.5021448158082508}}
-:::MLLOG {"namespace": "", "time_ms": 1745771201704, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745771218087, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745771218087, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6750143703960237, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6750143703960237}}
-:::MLLOG {"namespace": "", "time_ms": 1745772067913, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745772084343, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745772084343, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7019343614578247, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7019343614578247}}
-:::MLLOG {"namespace": "", "time_ms": 1745772937902, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745772954292, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745772954292, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7071203810828073, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7071203810828073}}
-:::MLLOG {"namespace": "", "time_ms": 1745773806065, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745773822470, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745773822470, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7098041250592186, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7098041250592186}}
-:::MLLOG {"namespace": "", "time_ms": 1745774672651, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745774689034, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745774689034, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7114683633758908, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7114683633758908}}
-:::MLLOG {"namespace": "", "time_ms": 1745775540181, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745775556554, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745775556554, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7130643929753985, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7130643929753985}}
-:::MLLOG {"namespace": "", "time_ms": 1745776408843, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745776425260, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745776425260, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135605426061721, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7135605426061721}}
-:::MLLOG {"namespace": "", "time_ms": 1745777274643, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745777291080, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745777291080, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7143973344848269, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7143973344848269}}
-:::MLLOG {"namespace": "", "time_ms": 1745778145622, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745778161980, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745778161980, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7153716751507351, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7153716751507351}}
-:::MLLOG {"namespace": "", "time_ms": 1745779016330, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745779032715, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745779032715, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7156423943383353, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7156423943383353}}
-:::MLLOG {"namespace": "", "time_ms": 1745779880722, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745779897142, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745779897142, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7163417912664868, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7163417912664868}}
-:::MLLOG {"namespace": "", "time_ms": 1745780753157, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745780769553, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745780769554, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7172225815909249, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7172225815909249}}
-:::MLLOG {"namespace": "", "time_ms": 1745781622729, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745781639078, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745781639078, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7175244842256818, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7175244842256818}}
-:::MLLOG {"namespace": "", "time_ms": 1745782486272, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745782502649, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745782502649, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186253524961925, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7186253524961925}}
-:::MLLOG {"namespace": "", "time_ms": 1745783352552, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745783368943, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745783368944, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186818900562468, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7186818900562468}}
-:::MLLOG {"namespace": "", "time_ms": 1745784226008, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745784242458, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745784242459, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7187245652789161, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7187245652789161}}
-:::MLLOG {"namespace": "", "time_ms": 1745785100858, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745785117223, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745785117223, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719807653767722, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.719807653767722}}
-:::MLLOG {"namespace": "", "time_ms": 1745785971577, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745785987960, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745785987960, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7201530842554001, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7201530842554001}}
-:::MLLOG {"namespace": "", "time_ms": 1745785987961, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3148992}}
-:::MLLOG {"namespace": "", "time_ms": 1745785987961, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745786017497, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017512, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017512, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017512, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017512, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017635, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745786017635, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745787211109, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745787229721, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745787229735, "event_type": "POINT_IN_TIME", "key": "seed", "value": 32670, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245941, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245941, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245942, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245943, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745787245944, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745787291328, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745788420396, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745788465722, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745788465723, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3736442960443951, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3736442960443951}}
-:::MLLOG {"namespace": "", "time_ms": 1745789321979, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745789338432, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745789338433, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.400805332830974, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.400805332830974}}
-:::MLLOG {"namespace": "", "time_ms": 1745790194764, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745790211239, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745790211239, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.46333109367461434, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.46333109367461434}}
-:::MLLOG {"namespace": "", "time_ms": 1745791069993, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745791086453, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745791086454, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.606940507888794, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.606940507888794}}
-:::MLLOG {"namespace": "", "time_ms": 1745791944536, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745791961053, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745791961053, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6969162089484079, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6969162089484079}}
-:::MLLOG {"namespace": "", "time_ms": 1745792823274, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745792839814, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745792839815, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7066007188388279, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7066007188388279}}
-:::MLLOG {"namespace": "", "time_ms": 1745793701391, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745793717919, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745793717919, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7095723350842794, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7095723350842794}}
-:::MLLOG {"namespace": "", "time_ms": 1745794576972, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745794593484, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745794593485, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7117142688660395, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7117142688660395}}
-:::MLLOG {"namespace": "", "time_ms": 1745795452646, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745795469118, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745795469119, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7129210210981823, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7129210210981823}}
-:::MLLOG {"namespace": "", "time_ms": 1745796328189, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745796346080, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745796346081, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7135851048287891, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7135851048287891}}
-:::MLLOG {"namespace": "", "time_ms": 1745797206117, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745797222639, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745797222639, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7144846972965059, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7144846972965059}}
-:::MLLOG {"namespace": "", "time_ms": 1745798099453, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745798115967, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745798115967, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7151809629939851, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7151809629939851}}
-:::MLLOG {"namespace": "", "time_ms": 1745798972875, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745798989390, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745798989390, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.715795217809223, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.715795217809223}}
-:::MLLOG {"namespace": "", "time_ms": 1745799851968, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745799868449, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745799868449, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7165858081408909, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7165858081408909}}
-:::MLLOG {"namespace": "", "time_ms": 1745800726576, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745800743139, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745800743139, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7168737269583203, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7168737269583203}}
-:::MLLOG {"namespace": "", "time_ms": 1745801600341, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745801616830, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745801616830, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7174848630314782, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7174848630314782}}
-:::MLLOG {"namespace": "", "time_ms": 1745802471269, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745802487752, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745802487753, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178929272152129, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7178929272152129}}
-:::MLLOG {"namespace": "", "time_ms": 1745803350240, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745803366751, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745803366751, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7188977406138466, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7188977406138466}}
-:::MLLOG {"namespace": "", "time_ms": 1745804227786, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745804245700, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745804245700, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190916072754633, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7190916072754633}}
-:::MLLOG {"namespace": "", "time_ms": 1745805107689, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745805124208, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745805124208, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7195343852043152, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7195343852043152}}
-:::MLLOG {"namespace": "", "time_ms": 1745805981361, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745805997841, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745805997842, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7201732056481498, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7201732056481498}}
-:::MLLOG {"namespace": "", "time_ms": 1745805997842, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3148992}}
-:::MLLOG {"namespace": "", "time_ms": 1745805997842, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,84 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745806025363, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025378, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025378, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025378, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025378, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025507, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745806025507, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745807261586, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745807280095, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745807280110, "event_type": "POINT_IN_TIME", "key": "seed", "value": 25877, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296270, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296271, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296271, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296271, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296271, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296271, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296272, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296272, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296272, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296272, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296272, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745807296273, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745807345166, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745808458376, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745808503186, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745808503187, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.37366413615998767, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.37366413615998767}}
-:::MLLOG {"namespace": "", "time_ms": 1745809355092, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745809371487, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745809371488, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.41079950133959453, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.41079950133959453}}
-:::MLLOG {"namespace": "", "time_ms": 1745810222251, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745810238608, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745810238609, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6318097074826559, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.6318097074826559}}
-:::MLLOG {"namespace": "", "time_ms": 1745811087580, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745811103948, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745811103948, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6998206564358302, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6998206564358302}}
-:::MLLOG {"namespace": "", "time_ms": 1745811956425, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745811972824, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745811972825, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7077281526156834, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7077281526156834}}
-:::MLLOG {"namespace": "", "time_ms": 1745812823492, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745812841248, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745812841248, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7099457195826939, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7099457195826939}}
-:::MLLOG {"namespace": "", "time_ms": 1745813692988, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745813709364, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745813709364, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7113831173805963, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7113831173805963}}
-:::MLLOG {"namespace": "", "time_ms": 1745814562349, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745814578741, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745814578741, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7134005540893191, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7134005540893191}}
-:::MLLOG {"namespace": "", "time_ms": 1745815431564, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745815447942, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745815447942, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7145692609605335, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7145692609605335}}
-:::MLLOG {"namespace": "", "time_ms": 1745816298939, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745816315317, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745816315317, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7152422836848668, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7152422836848668}}
-:::MLLOG {"namespace": "", "time_ms": 1745817166658, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745817183040, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745817183040, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7160442278498695, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7160442278498695}}
-:::MLLOG {"namespace": "", "time_ms": 1745818032767, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745818050528, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745818050529, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7162078914188204, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7162078914188204}}
-:::MLLOG {"namespace": "", "time_ms": 1745818901002, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745818917373, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745818917374, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717176822253636, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.717176822253636}}
-:::MLLOG {"namespace": "", "time_ms": 1745819785490, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745819801825, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745819801826, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7179754376411438, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7179754376411438}}
-:::MLLOG {"namespace": "", "time_ms": 1745820657021, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745820673391, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745820673392, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7183612340972537, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7183612340972537}}
-:::MLLOG {"namespace": "", "time_ms": 1745821521217, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745821537581, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745821537581, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189742480005537, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7189742480005537}}
-:::MLLOG {"namespace": "", "time_ms": 1745822387221, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745822403579, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745822403579, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7194264003208706, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7194264003208706}}
-:::MLLOG {"namespace": "", "time_ms": 1745823259046, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745823275427, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745823275427, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200029481024969, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7200029481024969}}
-:::MLLOG {"namespace": "", "time_ms": 1745823275427, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2699136}}
-:::MLLOG {"namespace": "", "time_ms": 1745823275427, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,93 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745823302673, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302687, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302688, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302688, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302688, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302819, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745823302819, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745824543991, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745824562214, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745824562229, "event_type": "POINT_IN_TIME", "key": "seed", "value": 28100, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578735, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578735, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578735, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578735, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578735, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578736, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578737, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578737, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578737, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578737, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745824578737, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745824627870, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745825742554, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745825790935, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745825790935, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3742778738339742, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3742778738339742}}
-:::MLLOG {"namespace": "", "time_ms": 1745826639341, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745826655849, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745826655850, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.41123455422265187, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.41123455422265187}}
-:::MLLOG {"namespace": "", "time_ms": 1745827506044, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745827522610, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745827522611, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.49199818543025425, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.49199818543025425}}
-:::MLLOG {"namespace": "", "time_ms": 1745828373508, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745828390052, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745828390052, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6114951428912935, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6114951428912935}}
-:::MLLOG {"namespace": "", "time_ms": 1745829241326, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745829257895, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745829257896, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7013765476998829, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7013765476998829}}
-:::MLLOG {"namespace": "", "time_ms": 1745830111499, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745830128070, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745830128070, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7078263350895473, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7078263350895473}}
-:::MLLOG {"namespace": "", "time_ms": 1745830981253, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745830997843, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745830997843, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7105506675583976, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7105506675583976}}
-:::MLLOG {"namespace": "", "time_ms": 1745831850249, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745831866851, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745831866851, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7127624767167228, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7127624767167228}}
-:::MLLOG {"namespace": "", "time_ms": 1745832720859, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745832737429, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745832737429, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7134371825626918, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7134371825626918}}
-:::MLLOG {"namespace": "", "time_ms": 1745833590656, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745833607244, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745833607244, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7143038937023708, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7143038937023708}}
-:::MLLOG {"namespace": "", "time_ms": 1745834459648, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745834476236, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745834476236, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.715397086029961, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.715397086029961}}
-:::MLLOG {"namespace": "", "time_ms": 1745835329853, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745835346430, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745835346430, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7157154185431344, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7157154185431344}}
-:::MLLOG {"namespace": "", "time_ms": 1745836205647, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745836222255, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745836222255, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7164555277143206, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7164555277143206}}
-:::MLLOG {"namespace": "", "time_ms": 1745837074549, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745837091187, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745837091187, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7164888756615775, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7164888756615775}}
-:::MLLOG {"namespace": "", "time_ms": 1745837955933, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745837972531, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745837972531, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7169959885733468, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7169959885733468}}
-:::MLLOG {"namespace": "", "time_ms": 1745838824798, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745838841423, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745838841424, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717557556288583, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.717557556288583}}
-:::MLLOG {"namespace": "", "time_ms": 1745839692680, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745839709268, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745839709269, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7181111250604902, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7181111250604902}}
-:::MLLOG {"namespace": "", "time_ms": 1745840566476, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745840583081, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745840583082, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189842644191924, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7189842644191924}}
-:::MLLOG {"namespace": "", "time_ms": 1745841439990, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745841456632, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745841456632, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190832745461236, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7190832745461236}}
-:::MLLOG {"namespace": "", "time_ms": 1745842320481, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745842337097, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745842337097, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192168309575036, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7192168309575036}}
-:::MLLOG {"namespace": "", "time_ms": 1745843193503, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745843210131, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745843210131, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7202679208346776, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7202679208346776}}
-:::MLLOG {"namespace": "", "time_ms": 1745843210131, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3148992}}
-:::MLLOG {"namespace": "", "time_ms": 1745843210131, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,102 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745787802566, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802594, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802594, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802594, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802594, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802726, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745787802727, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745789594073, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745789613521, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745789613548, "event_type": "POINT_IN_TIME", "key": "seed", "value": 19308, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635762, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635763, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635763, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635763, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635763, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635763, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635764, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635764, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635764, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635764, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635764, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635765, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635765, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635765, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635765, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635766, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745789635766, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745789698057, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745790966288, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745791025068, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745791025069, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3743333379427592, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3743333379427592}}
-:::MLLOG {"namespace": "", "time_ms": 1745791894750, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745791911957, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745791911958, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.40425296141987754, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.40425296141987754}}
-:::MLLOG {"namespace": "", "time_ms": 1745792783857, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745792801077, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745792801077, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.492629117057437, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.492629117057437}}
-:::MLLOG {"namespace": "", "time_ms": 1745793672991, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745793690266, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745793690267, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6120388820057824, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.6120388820057824}}
-:::MLLOG {"namespace": "", "time_ms": 1745794563811, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745794581118, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745794581119, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6954671644029163, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6954671644029163}}
-:::MLLOG {"namespace": "", "time_ms": 1745795456868, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745795474160, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745795474160, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7052601121720814, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7052601121720814}}
-:::MLLOG {"namespace": "", "time_ms": 1745796348727, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745796365991, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745796365992, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7080547185171218, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7080547185171218}}
-:::MLLOG {"namespace": "", "time_ms": 1745797241370, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745797258609, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745797258609, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7103598571958996, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7103598571958996}}
-:::MLLOG {"namespace": "", "time_ms": 1745798133476, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745798150737, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745798150737, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7112415722438268, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7112415722438268}}
-:::MLLOG {"namespace": "", "time_ms": 1745799023898, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745799041165, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745799041165, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7124293872288295, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7124293872288295}}
-:::MLLOG {"namespace": "", "time_ms": 1745799924843, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745799942107, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745799942107, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7139696609406244, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7139696609406244}}
-:::MLLOG {"namespace": "", "time_ms": 1745800815328, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745800832615, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745800832615, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7143780969438098, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7143780969438098}}
-:::MLLOG {"namespace": "", "time_ms": 1745801712120, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745801729406, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745801729406, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7155611310686384, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7155611310686384}}
-:::MLLOG {"namespace": "", "time_ms": 1745802601680, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745802618955, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745802618956, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7154839975493295, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7154839975493295}}
-:::MLLOG {"namespace": "", "time_ms": 1745803490060, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745803507331, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745803507332, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716192881266276, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.716192881266276}}
-:::MLLOG {"namespace": "", "time_ms": 1745804380068, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745804397307, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745804397308, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7166955840019953, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7166955840019953}}
-:::MLLOG {"namespace": "", "time_ms": 1745805278057, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745805295302, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745805295303, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717407656851269, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.717407656851269}}
-:::MLLOG {"namespace": "", "time_ms": 1745806170941, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745806188188, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745806188188, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7178891250065395, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7178891250065395}}
-:::MLLOG {"namespace": "", "time_ms": 1745807059891, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745807077172, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745807077172, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7180268843968709, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7180268843968709}}
-:::MLLOG {"namespace": "", "time_ms": 1745807956172, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745807973421, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745807973422, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7189508608409336, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.7189508608409336}}
-:::MLLOG {"namespace": "", "time_ms": 1745808865958, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3148992, "step_num": 32802}}
-:::MLLOG {"namespace": "", "time_ms": 1745808883154, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3148992, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3148992, "step_num": 32802, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745808883154, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192203856649853, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3148992, "masked_lm_accuracy": 0.7192203856649853}}
-:::MLLOG {"namespace": "", "time_ms": 1745809754328, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3298944, "step_num": 34364}}
-:::MLLOG {"namespace": "", "time_ms": 1745809771550, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3298944, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3298944, "step_num": 34364, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745809771550, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.719561649504162, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3298944, "masked_lm_accuracy": 0.719561649504162}}
-:::MLLOG {"namespace": "", "time_ms": 1745810642103, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3448896, "step_num": 35926}}
-:::MLLOG {"namespace": "", "time_ms": 1745810659332, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3448896, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3448896, "step_num": 35926, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745810659333, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7198069998196193, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3448896, "masked_lm_accuracy": 0.7198069998196193}}
-:::MLLOG {"namespace": "", "time_ms": 1745811532161, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 3598848, "step_num": 37488}}
-:::MLLOG {"namespace": "", "time_ms": 1745811549386, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 3598848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 3598848, "step_num": 37488, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745811549386, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7200410479591006, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 3598848, "masked_lm_accuracy": 0.7200410479591006}}
-:::MLLOG {"namespace": "", "time_ms": 1745811549387, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 3598848, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 3598848}}
-:::MLLOG {"namespace": "", "time_ms": 1745811549387, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,90 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745811581618, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581646, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581646, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581646, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581646, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581770, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745811581771, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745813333105, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745813352889, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745813412916, "event_type": "POINT_IN_TIME", "key": "seed", "value": 383, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435199, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435199, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435200, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435200, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435200, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435200, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435201, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435201, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435201, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435201, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435202, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435202, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435202, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435202, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435203, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435203, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745813435203, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745813498099, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745814758540, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745814817661, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745814817662, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3754178938411531, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3754178938411531}}
-:::MLLOG {"namespace": "", "time_ms": 1745815683185, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745815700873, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745815700874, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4043772734346844, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4043772734346844}}
-:::MLLOG {"namespace": "", "time_ms": 1745816568437, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745816586125, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745816586125, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.46416939128012885, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.46416939128012885}}
-:::MLLOG {"namespace": "", "time_ms": 1745817454256, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745817471985, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745817471985, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5808823159762792, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.5808823159762792}}
-:::MLLOG {"namespace": "", "time_ms": 1745818340940, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745818358717, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745818358718, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.6975844576245263, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.6975844576245263}}
-:::MLLOG {"namespace": "", "time_ms": 1745819230249, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745819247985, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745819247985, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7072370296432858, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7072370296432858}}
-:::MLLOG {"namespace": "", "time_ms": 1745820119013, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745820136733, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745820136733, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7102725159554254, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.7102725159554254}}
-:::MLLOG {"namespace": "", "time_ms": 1745821006045, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745821023778, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745821023779, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7125218317622231, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7125218317622231}}
-:::MLLOG {"namespace": "", "time_ms": 1745821894699, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745821912431, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745821912431, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7138815391631353, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7138815391631353}}
-:::MLLOG {"namespace": "", "time_ms": 1745822784454, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745822803745, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745822803746, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.714685621148064, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.714685621148064}}
-:::MLLOG {"namespace": "", "time_ms": 1745823676721, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745823694574, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745823694575, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7150295643579392, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7150295643579392}}
-:::MLLOG {"namespace": "", "time_ms": 1745824569435, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745824587278, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745824587278, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7158357126372201, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7158357126372201}}
-:::MLLOG {"namespace": "", "time_ms": 1745825476604, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745825494462, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745825494462, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.716386129742577, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.716386129742577}}
-:::MLLOG {"namespace": "", "time_ms": 1745826367004, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745826384822, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745826384822, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7171017362957909, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7171017362957909}}
-:::MLLOG {"namespace": "", "time_ms": 1745827257763, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745827275559, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745827275559, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.717421099117824, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.717421099117824}}
-:::MLLOG {"namespace": "", "time_ms": 1745828152818, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745828170593, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745828170594, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7179956731342134, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7179956731342134}}
-:::MLLOG {"namespace": "", "time_ms": 1745829042290, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2549184, "step_num": 26554}}
-:::MLLOG {"namespace": "", "time_ms": 1745829060104, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2549184, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2549184, "step_num": 26554, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745829060105, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7186045289039612, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2549184, "masked_lm_accuracy": 0.7186045289039612}}
-:::MLLOG {"namespace": "", "time_ms": 1745829932500, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2699136, "step_num": 28116}}
-:::MLLOG {"namespace": "", "time_ms": 1745829950294, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2699136, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2699136, "step_num": 28116, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745829950294, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190587202707927, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2699136, "masked_lm_accuracy": 0.7190587202707927}}
-:::MLLOG {"namespace": "", "time_ms": 1745830838162, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2849088, "step_num": 29678}}
-:::MLLOG {"namespace": "", "time_ms": 1745830855978, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2849088, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2849088, "step_num": 29678, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745830855978, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7192222112701052, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2849088, "masked_lm_accuracy": 0.7192222112701052}}
-:::MLLOG {"namespace": "", "time_ms": 1745831740334, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2999040, "step_num": 31240}}
-:::MLLOG {"namespace": "", "time_ms": 1745831758212, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2999040, "step_num": 31240, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745831758213, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.720143475418999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2999040, "masked_lm_accuracy": 0.720143475418999}}
-:::MLLOG {"namespace": "", "time_ms": 1745831758213, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2999040, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2999040}}
-:::MLLOG {"namespace": "", "time_ms": 1745831758213, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,78 +0,0 @@
-:::MLLOG {"namespace": "", "time_ms": 1745831791121, "event_type": "POINT_IN_TIME", "key": "submission_org", "value": "tinycorp", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 917}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791148, "event_type": "POINT_IN_TIME", "key": "submission_platform", "value": "tinybox_red", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 918}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791148, "event_type": "POINT_IN_TIME", "key": "submission_division", "value": "closed", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 919}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791148, "event_type": "POINT_IN_TIME", "key": "submission_status", "value": "onprem", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 920}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791148, "event_type": "POINT_IN_TIME", "key": "submission_benchmark", "value": "bert", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 922}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791273, "event_type": "POINT_IN_TIME", "key": "cache_clear", "value": true, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 925}}
-:::MLLOG {"namespace": "", "time_ms": 1745831791274, "event_type": "INTERVAL_START", "key": "init_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 926}}
-:::MLLOG {"namespace": "", "time_ms": 1745833591107, "event_type": "POINT_IN_TIME", "key": "init_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1138}}
-:::MLLOG {"namespace": "", "time_ms": 1745833610931, "event_type": "INTERVAL_START", "key": "run_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 929}}
-:::MLLOG {"namespace": "", "time_ms": 1745833640959, "event_type": "POINT_IN_TIME", "key": "seed", "value": 10542, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 930}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663181, "event_type": "POINT_IN_TIME", "key": "global_batch_size", "value": 96, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1007}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663182, "event_type": "POINT_IN_TIME", "key": "max_sequence_length", "value": 512, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1008}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663182, "event_type": "POINT_IN_TIME", "key": "max_predictions_per_seq", "value": 76, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1009}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663182, "event_type": "POINT_IN_TIME", "key": "opt_name", "value": "LAMB", "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1011}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663183, "event_type": "POINT_IN_TIME", "key": "opt_base_learning_rate", "value": 0.000175, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1012}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663183, "event_type": "POINT_IN_TIME", "key": "opt_lamb_weight_decay_rate", "value": 0.01, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1013}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663183, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_1", "value": 0.9, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1014}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663183, "event_type": "POINT_IN_TIME", "key": "opt_lamb_beta_2", "value": 0.999, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1015}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663184, "event_type": "POINT_IN_TIME", "key": "opt_lamb_learning_rate_decay_poly_power", "value": 1.0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1016}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663184, "event_type": "POINT_IN_TIME", "key": "opt_lamb_epsilon", "value": 1e-06, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1017}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663184, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1019}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663184, "event_type": "POINT_IN_TIME", "key": "num_warmup_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1020}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663184, "event_type": "POINT_IN_TIME", "key": "start_warmup_step", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1021}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663185, "event_type": "POINT_IN_TIME", "key": "opt_learning_rate_training_steps", "value": 37500, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1022}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663185, "event_type": "POINT_IN_TIME", "key": "gradient_accumulation_steps", "value": 1, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1023}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663185, "event_type": "POINT_IN_TIME", "key": "eval_samples", "value": 10080, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1024}}
-:::MLLOG {"namespace": "", "time_ms": 1745833663185, "event_type": "POINT_IN_TIME", "key": "train_samples", "value": 3600000, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1025}}
-:::MLLOG {"namespace": "", "time_ms": 1745833717917, "event_type": "INTERVAL_START", "key": "epoch_start", "value": 0, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1055, "epoch_num": 0}}
-:::MLLOG {"namespace": "", "time_ms": 1745834985679, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 149952, "step_num": 1562}}
-:::MLLOG {"namespace": "", "time_ms": 1745835045156, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 149952, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 149952, "step_num": 1562, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745835045156, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.3788086683977218, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 149952, "masked_lm_accuracy": 0.3788086683977218}}
-:::MLLOG {"namespace": "", "time_ms": 1745835918374, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 299904, "step_num": 3124}}
-:::MLLOG {"namespace": "", "time_ms": 1745835935660, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 299904, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 299904, "step_num": 3124, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745835935660, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.4123597074122656, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 299904, "masked_lm_accuracy": 0.4123597074122656}}
-:::MLLOG {"namespace": "", "time_ms": 1745836810436, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 449856, "step_num": 4686}}
-:::MLLOG {"namespace": "", "time_ms": 1745836827731, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 449856, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 449856, "step_num": 4686, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745836827731, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.5868216173989432, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 449856, "masked_lm_accuracy": 0.5868216173989432}}
-:::MLLOG {"namespace": "", "time_ms": 1745837700592, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 599808, "step_num": 6248}}
-:::MLLOG {"namespace": "", "time_ms": 1745837719403, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 599808, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 599808, "step_num": 6248, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745837719404, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7020311525889805, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 599808, "masked_lm_accuracy": 0.7020311525889805}}
-:::MLLOG {"namespace": "", "time_ms": 1745838594733, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 749760, "step_num": 7810}}
-:::MLLOG {"namespace": "", "time_ms": 1745838612057, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 749760, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 749760, "step_num": 7810, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745838612058, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7091197768847147, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 749760, "masked_lm_accuracy": 0.7091197768847147}}
-:::MLLOG {"namespace": "", "time_ms": 1745839489790, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 899712, "step_num": 9372}}
-:::MLLOG {"namespace": "", "time_ms": 1745839507098, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 899712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 899712, "step_num": 9372, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745839507098, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7120704963093712, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 899712, "masked_lm_accuracy": 0.7120704963093712}}
-:::MLLOG {"namespace": "", "time_ms": 1745840383347, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1049664, "step_num": 10934}}
-:::MLLOG {"namespace": "", "time_ms": 1745840400642, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1049664, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1049664, "step_num": 10934, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745840400643, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.713741013549623, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1049664, "masked_lm_accuracy": 0.713741013549623}}
-:::MLLOG {"namespace": "", "time_ms": 1745841277938, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1199616, "step_num": 12496}}
-:::MLLOG {"namespace": "", "time_ms": 1745841295237, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1199616, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1199616, "step_num": 12496, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745841295238, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7150049794287908, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1199616, "masked_lm_accuracy": 0.7150049794287908}}
-:::MLLOG {"namespace": "", "time_ms": 1745842172781, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1349568, "step_num": 14058}}
-:::MLLOG {"namespace": "", "time_ms": 1745842190096, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1349568, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1349568, "step_num": 14058, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745842190096, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7166263137544905, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1349568, "masked_lm_accuracy": 0.7166263137544905}}
-:::MLLOG {"namespace": "", "time_ms": 1745843065811, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1499520, "step_num": 15620}}
-:::MLLOG {"namespace": "", "time_ms": 1745843083177, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1499520, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1499520, "step_num": 15620, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745843083177, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7167705734570821, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1499520, "masked_lm_accuracy": 0.7167705734570821}}
-:::MLLOG {"namespace": "", "time_ms": 1745843967823, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1649472, "step_num": 17182}}
-:::MLLOG {"namespace": "", "time_ms": 1745843985219, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1649472, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1649472, "step_num": 17182, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745843985219, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7174085940633501, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1649472, "masked_lm_accuracy": 0.7174085940633501}}
-:::MLLOG {"namespace": "", "time_ms": 1745844871164, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1799424, "step_num": 18744}}
-:::MLLOG {"namespace": "", "time_ms": 1745844888581, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1799424, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1799424, "step_num": 18744, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745844888582, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7175841513134185, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1799424, "masked_lm_accuracy": 0.7175841513134185}}
-:::MLLOG {"namespace": "", "time_ms": 1745845771506, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 1949376, "step_num": 20306}}
-:::MLLOG {"namespace": "", "time_ms": 1745845788927, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 1949376, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 1949376, "step_num": 20306, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745845788928, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7190523573330471, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 1949376, "masked_lm_accuracy": 0.7190523573330471}}
-:::MLLOG {"namespace": "", "time_ms": 1745846665765, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2099328, "step_num": 21868}}
-:::MLLOG {"namespace": "", "time_ms": 1745846683091, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2099328, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2099328, "step_num": 21868, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745846683091, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7195040078390212, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2099328, "masked_lm_accuracy": 0.7195040078390212}}
-:::MLLOG {"namespace": "", "time_ms": 1745847564414, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2249280, "step_num": 23430}}
-:::MLLOG {"namespace": "", "time_ms": 1745847581796, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2249280, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2249280, "step_num": 23430, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745847581797, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7199954066957747, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2249280, "masked_lm_accuracy": 0.7199954066957747}}
-:::MLLOG {"namespace": "", "time_ms": 1745848457298, "event_type": "INTERVAL_START", "key": "eval_start", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1106, "epoch_num": 2399232, "step_num": 24992}}
-:::MLLOG {"namespace": "", "time_ms": 1745848474654, "event_type": "INTERVAL_END", "key": "eval_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1159, "epoch_count": 2399232, "step_num": 24992, "samples_count": 10080}}
-:::MLLOG {"namespace": "", "time_ms": 1745848474654, "event_type": "POINT_IN_TIME", "key": "eval_accuracy", "value": 0.7207992462884812, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1160, "epoch_num": 2399232, "masked_lm_accuracy": 0.7207992462884812}}
-:::MLLOG {"namespace": "", "time_ms": 1745848474655, "event_type": "POINT_IN_TIME", "key": "epoch_stop", "value": 2399232, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1178, "epoch_num": 2399232}}
-:::MLLOG {"namespace": "", "time_ms": 1745848474655, "event_type": "INTERVAL_END", "key": "run_stop", "value": null, "metadata": {"file": "tinygrad/examples/mlperf/model_train.py", "lineno": 1179, "status": "success"}}
@@ -1,39 +0,0 @@
-{
-    "submitter": "tinycorp",
-    "division": "closed",
-    "status": "Available on-premise",
-    "system_name": "tinybox 8xMI300X",
-    "number_of_nodes": "1",
-    "host_processors_per_node": "2",
-    "host_processor_model_name": "AMD EPYC 9354 32-Core Processor",
-    "host_processor_core_count": "32",
-    "host_processor_vcpu_count": "64",
-    "host_processor_frequency": "",
-    "host_processor_caches": "",
-    "host_processor_interconnect": "",
-    "host_memory_capacity": "2304GB",
-    "host_storage_type": "NVMe SSD",
-    "host_storage_capacity": "3x 4TB raid array",
-    "host_networking": "",
-    "host_networking_topology": "",
-    "host_memory_configuration": "24x 96GB DDR5",
-    "accelerators_per_node": "8",
-    "accelerator_model_name": "AMD Instinct MI300X",
-    "accelerator_host_interconnect": "PCIe 5.0 x16",
-    "accelerator_frequency": "",
-    "accelerator_on-chip_memories": "",
-    "accelerator_memory_configuration": "HBM3",
-    "accelerator_memory_capacity": "192GB",
-    "accelerator_interconnect": "",
-    "accelerator_interconnect_topology": "",
-    "cooling": "air",
-    "hw_notes": "",
-    "framework": "tinygrad, branch mlperf_training_v5.0",
-    "other_software_stack": {
-      "python": "3.10.16",
-      "ROCm": "3.0.0+94441cb"
-    },
-    "operating_system": "Ubuntu 24.04.1 LTS",
-    "sw_notes": ""
-  }
-  
@@ -28,7 +28,7 @@
  "accelerator_interconnect_topology": "",
  "cooling": "air",
  "hw_notes": "",
-  "framework": "tinygrad, branch mlperf_training_v5.0",
+  "framework": "tinygrad, commit b5546912e24e0a864b35924da4efa5d71cfe368b",
  "other_software_stack": {
    "python": "3.10.12",
    "CUDA": "12.4"
@@ -28,9 +28,10 @@
  "accelerator_interconnect_topology": "",
  "cooling": "air",
  "hw_notes": "",
-  "framework": "tinygrad, branch mlperf_training_v5.0",
+  "framework": "tinygrad, commit b5546912e24e0a864b35924da4efa5d71cfe368b",
  "other_software_stack": {
-    "python": "3.10.12"
+    "python": "3.10.12",
+    "ROCm": "6.1.3"
  },
  "operating_system": "Ubuntu 22.04.4",
  "sw_notes": ""
@@ -1,11 +1,10 @@
 # https://arxiv.org/pdf/2409.02060
-import time
 import numpy as np
 np.set_printoptions(suppress=True, linewidth=1000)
-import functools
-from tinygrad import Tensor, nn, Device, GlobalCounters
-from tinygrad.helpers import Timing, getenv
-from extra.models.llama import Transformer, convert_from_huggingface
+import functools, collections, json
+from tinygrad import Tensor, nn, Device
+from tinygrad.helpers import tqdm, CI, Profiling, Timing, fetch, getenv
+from extra.models.llama import Transformer, Variable, convert_from_huggingface

 class MixtureFeedForward:
  def __init__(self, num_experts:int, activated_experts:int, dim:int, hidden_dim:int, linear=nn.Linear):
@@ -19,8 +18,11 @@ class MixtureFeedForward:
    assert x.shape[1] == 1, "only length=1"
    g = self.gate(x).float().softmax(-1)

-    g = g.squeeze() # (BS, length, num_experts) -> (num_experts,)
-    probs, sel = g.topk(self.activated_experts)
+    # TODO: don't go to CPU here
+    choice = g.data().tolist()[0][0]
+    top = sorted(enumerate(choice), key=lambda x: -x[1])[:self.activated_experts]
+    sel, probs = Tensor([x[0] for x in top]), Tensor([x[1] for x in top])
+    #print(sel.numpy(), probs.numpy())

    # run MoE
    x_up_gate = x.dot(self.gate_proj[sel].permute(0,2,1)).silu() * x.dot(self.up_proj[sel].permute(0,2,1))
@@ -50,21 +52,17 @@ if __name__ == "__main__":

  with Timing("create model: "):
    model = Transformer(n_layers=16, dim=2048, hidden_dim=1024, n_heads=16, norm_eps=1e-5, qk_norm=1e-5, max_context=1024,
-                        vocab_size=50304, feed_forward=functools.partial(MixtureFeedForward, 64, 8))
+                        vocab_size=50304, feed_forward=functools.partial(MixtureFeedForward, 64, 8), jit=False)
    model_state_dict = nn.state.get_state_dict(model)
    del model_state_dict['freqs_cis']

-  with Timing("load weights to GPU: "):
-    nhf_state = convert_from_huggingface(fetch_weights(), model, 16, 16)
+  with Timing("fetch and load weights: "):
+    state = fetch_weights()
+    nhf_state = convert_from_huggingface(state, model, 16, 16)
    # NOTE: i'm not sure this actually needs float32, it may just change the type of things downstream from it. but doesn't match torch w/o this
    for needs_float32 in ['tok_embeddings.weight']: nhf_state[needs_float32] = nhf_state[needs_float32].float()
-  print(f"ram used: {GlobalCounters.mem_used/1e9:.2f} GB")
-
-  with Timing("unpack weights: "):
    nn.state.load_state_dict(model, nhf_state, verbose=False, strict=False, consume=True, realize=False)
    assert len(nhf_state) == 0
-    Tensor.realize(*list(nn.state.get_state_dict(model).values()))
-  print(f"ram used: {GlobalCounters.mem_used/1e9:.2f} GB")

  count = 30
  temperature = 0
@@ -75,20 +73,14 @@ if __name__ == "__main__":

  toks = [12092]
  start_pos = 0
-  timings = []
  for i in range(count):
-    GlobalCounters.reset()
-    st = time.perf_counter()
-    tok = model(Tensor([toks[start_pos:]]), start_pos, temperature).item()
-    timings.append(time.perf_counter()-st)
+    tok = model(Tensor([toks[start_pos:]]), 0 if start_pos == 0 else Variable("start_pos", 1, 1024).bind(start_pos), temperature).item()
    toks.append(tok)
    start_pos += 1
    print(toks)
    print(tokenizer.decode(toks))
-  print(f"fastest token {min(timings)*1e3:.2f} ms, {1/min(timings):.1f} tok/s")

-  if temperature == 0:
-    # Hello, I am a newbie to this forum and I am trying to get a better understanding of the different types of data that can be stored in a
-    assert toks == [12092, 13, 309, 717, 247, 747, 17782, 281, 436, 12209, 285, 309, 717, 2820, 281, 755,
-                    247, 1805, 4685, 273, 253, 1027, 3510, 273, 941, 326, 476, 320, 7141, 275, 247], "BAD OUTPUT!"
+  # Hello, I am a newbie to this forum and I am trying to get a better understanding of the different types of data that can be stored in a
+  assert toks == [12092, 13, 309, 717, 247, 747, 17782, 281, 436, 12209, 285, 309, 717, 2820, 281, 755,
+                  247, 1805, 4685, 273, 253, 1027, 3510, 273, 941, 326, 476, 320, 7141, 275, 247], "BAD OUTPUT!"

@@ -12,7 +12,7 @@ from tinygrad.engine.realize import CompiledRunner

 import onnx
 from onnx.helper import tensor_dtype_to_np_dtype
-from tinygrad.frontend.onnx import OnnxRunner
+from extra.onnx import OnnxRunner   # TODO: port to main tinygrad

 OPENPILOT_MODEL = sys.argv[1] if len(sys.argv) > 1 else "https://github.com/commaai/openpilot/raw/v0.9.7/selfdrive/modeld/models/supercombo.onnx"
 OUTPUT = sys.argv[2] if len(sys.argv) > 2 else "/tmp/openpilot.pkl"
@@ -1,5 +1,5 @@
-from tinygrad import dtypes, getenv, Device
-from tinygrad.helpers import trange, colored, DEBUG, temp
+from tinygrad import dtypes, getenv
+from tinygrad.helpers import trange, colored
 from tinygrad.nn.datasets import mnist
 import torch
 from torch import nn, optim
@@ -30,15 +30,13 @@ if __name__ == "__main__":
    import tinygrad.frontend.torch
    device = torch.device("tiny")
  else:
-    device = torch.device({"METAL":"mps","NV":"cuda"}.get(Device.DEFAULT, "cpu"))
-  if DEBUG >= 1: print(f"using torch backend {device}")
+    device = torch.device("mps")
  X_train, Y_train, X_test, Y_test = mnist()
  X_train = torch.tensor(X_train.float().numpy(), device=device)
  Y_train = torch.tensor(Y_train.cast(dtypes.int64).numpy(), device=device)
  X_test = torch.tensor(X_test.float().numpy(), device=device)
  Y_test = torch.tensor(Y_test.cast(dtypes.int64).numpy(), device=device)

-  if getenv("TORCHVIZ"): torch.cuda.memory._record_memory_history()
  model = Model().to(device)
  optimizer = optim.Adam(model.parameters(), 1e-3)

@@ -64,6 +62,3 @@ if __name__ == "__main__":
  if target := getenv("TARGET_EVAL_ACC_PCT", 0.0):
    if test_acc >= target and test_acc != 100.0: print(colored(f"{test_acc=} >= {target}", "green"))
    else: raise ValueError(colored(f"{test_acc=} < {target}", "red"))
-  if getenv("TORCHVIZ"):
-    torch.cuda.memory._dump_snapshot(fp:=temp("torchviz.pkl", append_user=True))
-    print(f"saved torch memory snapshot to {fp}, view in https://pytorch.org/memory_viz")
@@ -504,7 +504,7 @@ def load_checkpoint_enc(checkpoint_path, model: ContentVec, optimizer=None, skip
        obj, v = getattr(parent, "weight"), weight_norm(weight_v, weight_g, 0)
        weight_g, weight_v, parent, skip = None, None, None, False
      if not skip and obj.shape == v.shape:
-        if "feature_extractor" in key and (isinstance(parent, (nn.GroupNorm, nn.LayerNorm))):  # cast
+        if "feature_extractor" in key and (isinstance(parent, nn.GroupNorm) or isinstance(parent, nn.LayerNorm)):  # cast
          obj.assign(v.to(obj.device).float())
        else:
          obj.assign(v.to(obj.device))
@@ -58,7 +58,7 @@ if __name__ == "__main__":
            return None
          return {"input": img.numpy()}
      quantize_static(model_fp32, fn, ImagenetReader(), quant_format=QuantFormat.QDQ, per_channel=False,
-                      activation_type=QuantType.QUInt8, weight_type=QuantType.QUInt8,
+                      activation_type=QuantType.QUInt8, weight_type=QuantType.QInt8,
                      extra_options={"ActivationSymmetric": False})

  run_onnx_jit, input_specs = load_onnx_model(fetch(fn))
@@ -70,12 +70,9 @@ if __name__ == "__main__":
    GlobalCounters.reset()
    p = run_onnx_jit(**{t_name:img})
    assert p.shape == (1,1000)
-    t = p.to('cpu').argmax().item()
+    t = p.argmax().item()
    hit += y==t
    print(f"target: {y:3d}  pred: {t:3d}  acc: {hit/(i+1)*100:.2f}%")

-  MS_TARGET = 13.4
-  print(f"need {GlobalCounters.global_ops/1e9*(1000/MS_TARGET):.2f} GFLOPS for {MS_TARGET:.2f} ms")
-
  import pickle
  with open("/tmp/im.pkl", "wb") as f: pickle.dump(run_onnx_jit, f)
@@ -1,19 +0,0 @@
-import sys, pickle
-from tinygrad import GlobalCounters
-from tinygrad.helpers import fetch, getenv
-from examples.test_onnx_imagenet import imagenet_dataloader
-
-if __name__ == "__main__":
-  with open(fetch(sys.argv[1]), "rb") as f:
-    run_onnx_jit = pickle.load(f)
-  input_name = run_onnx_jit.captured.expected_names[0]
-  device = run_onnx_jit.captured.expected_st_vars_dtype_device[0][-1]
-  print(f"input goes into {input_name=} on {device=}")
-  hit = 0
-  for i,(img,y) in enumerate(imagenet_dataloader(cnt=getenv("CNT", 100))):
-    GlobalCounters.reset()
-    p = run_onnx_jit(**{input_name:img.to(device)})
-    assert p.shape == (1,1000)
-    t = p.to('cpu').argmax().item()
-    hit += y==t
-    print(f"target: {y:3d}  pred: {t:3d}  acc: {hit/(i+1)*100:.2f}%")
@@ -17,7 +17,7 @@ def make_dataset():
  random.shuffle(ds)
  ds = np.array(ds).astype(np.float32)
  ds_X = ds[:, 0:6]
-  ds_Y = np.copy(ds[:, 1:]).astype(np.int32)
+  ds_Y = np.copy(ds[:, 1:])
  ds_X_train, ds_X_test = ds_X[0:8000], ds_X[8000:]
  ds_Y_train, ds_Y_test = ds_Y[0:8000], ds_Y[8000:]
  return ds_X_train, ds_Y_train, ds_X_test, ds_Y_test
@@ -18,6 +18,10 @@ canvas {  display: none; }
 </style>
 <title>tinygrad has WebGPU</title>
 <link rel="icon" type="image/x-icon" href="https://raw.githubusercontent.com/tinygrad/tinygrad/master/docs/logo.png">
+<script type="module">
+	import model from "../../net.js";
+	window.model = model;
+</script>
 </head>
 <body>
 <h1>WebGPU <a href="https://github.com/geohot/tinygrad">tinygrad</a> EfficientNet!</h1>
@@ -98,7 +102,6 @@ canvas {  display: none; }
 			resultText.innerHTML = "loading..."
 			labels = await getLabels();
 			const device = await getDevice();
-			const model = (await import("../../net.js")).default;
 			net = await timer(() => model.load(device, '../../net.safetensors'), "(compilation)");
 			resultText.innerHTML = "ready"
 		} catch (e) {
@@ -3,7 +3,7 @@ import os
 from ultralytics import YOLO
 import onnx
 from pathlib import Path
-from tinygrad.frontend.onnx import OnnxRunner
+from extra.onnx import OnnxRunner
 from extra.onnx_helpers import get_example_inputs
 from tinygrad.tensor import Tensor

@@ -1,14 +1,14 @@
 #!/usr/bin/env python3

 import time, mmap, sys, shutil, os, glob, subprocess
-from tinygrad.helpers import DEBUG, colored, ansilen
+from tinygrad.helpers import to_mv, DEBUG, colored, ansilen
 from tinygrad.runtime.autogen import libc
-from tinygrad.runtime.autogen.am import am
-from tinygrad.runtime.support.hcq import MMIOInterface
+from tinygrad.runtime.autogen.am import smu_v13_0_0
 from tinygrad.runtime.support.am.amdev import AMDev, AMMemoryManager
 from tinygrad.runtime.support.am.ip import AM_SOC, AM_GMC, AM_IH, AM_PSP, AM_SMU, AM_GFX, AM_SDMA

 AM_VERSION = 0xA0000002
+SMU_11_0_GFX_BUSY_THRESHOLD = 15

 def bold(s): return f"\033[1m{s}\033[0m"

@@ -16,10 +16,6 @@ def trim(s:str, length:int) -> str:
  if len(s) > length: return s[:length-3] + "..."
  return s

-def pad(x:str, length:int) -> str:
-  if len(x) < length: return x + " " * (length - len(x))
-  return x
-
 def color_temp(temp):
  if temp >= 87: return colored(f"{temp:>3}", "red")
  elif temp >= 80: return colored(f"{temp:>3}", "yellow")
@@ -42,7 +38,7 @@ def same_line(strs:list[list[str]|None], split=8) -> list[str]:
  for i in range(max_height):
    line = []
    for bid, block in enumerate(strs):
-      if i < len(block): line.append(block[i] + (' ' * (split + max_width_in_block[bid] - ansilen(block[i])) if bid != len(strs) - 1 else ''))
+      if i < len(block): line.append(block[i] + ' ' * (split + max_width_in_block[bid] - ansilen(block[i])))
      else: line.append(' ' * (split + max_width_in_block[bid]))
    ret.append(' '.join(line))
  return ret
@@ -59,7 +55,7 @@ def get_bar0_size(pcibus):
  return int(end_hex, 16) - int(start_hex, 16) + 1

 class AMSMI(AMDev):
-  def __init__(self, pcibus, vram_bar:MMIOInterface, doorbell_bar:MMIOInterface, mmio_bar:MMIOInterface):
+  def __init__(self, pcibus, vram_bar:memoryview, doorbell_bar:memoryview, mmio_bar:memoryview):
    self.pcibus = pcibus
    self.vram, self.doorbell64, self.mmio = vram_bar, doorbell_bar, mmio_bar
    self.pci_state = self.read_pci_state()
@@ -83,8 +79,6 @@ class AMSMI(AMDev):
    self.psp:AM_PSP = AM_PSP(self)
    self.smu:AM_SMU = AM_SMU(self)

-    for ip in [self.soc, self.gmc, self.ih, self.psp, self.smu]: ip.init_sw()
-
  def read_pci_state(self):
    with open(f"/sys/bus/pci/devices/{self.pcibus}/power_state", "r") as f: return f.read().strip().rstrip()

@@ -107,9 +101,9 @@ class SMICtx:
      bar_fds = {bar: os.open(f"/sys/bus/pci/devices/{pcibus}/resource{bar}", os.O_RDWR | os.O_SYNC) for bar in [0, 2, 5]}
      bar_size = {0: get_bar0_size(pcibus), 2: os.fstat(bar_fds[2]).st_size, 5: os.fstat(bar_fds[5]).st_size}

-      def map_pci_range(bar, fmt='B'):
-        return MMIOInterface(libc.mmap(0, bar_size[bar], mmap.PROT_READ | mmap.PROT_WRITE, mmap.MAP_SHARED, bar_fds[bar], 0), bar_size[bar], fmt)
-      self.opened_pci_resources[pcibus] = (map_pci_range(0), None, map_pci_range(5, 'I'))
+      def map_pci_range(bar):
+        return to_mv(libc.mmap(0, bar_size[bar], mmap.PROT_READ | mmap.PROT_WRITE, mmap.MAP_SHARED, bar_fds[bar], 0), bar_size[bar])
+      self.opened_pci_resources[pcibus] = (map_pci_range(0), None, map_pci_range(5).cast('I'))

    try:
      self.devs.append(AMSMI(pcibus, *self.opened_pci_resources[pcibus]))
@@ -140,95 +134,83 @@ class SMICtx:

  def collect(self): return {d: d.smu.read_metrics() if d.pci_state == "D0" else None for d in self.devs}

-  def get_gfx_activity(self, dev, metrics): return metrics.SmuMetrics.AverageGfxActivity
-  def get_mem_activity(self, dev, metrics): return metrics.SmuMetrics.AverageUclkActivity
-
-  def get_temps(self, dev, metrics, compact=False):
-    temps_keys = [(k, name) for k, name in dev.smu.smu_mod.c__EA_TEMP_e__enumvalues.items()
-                  if k < dev.smu.smu_mod.TEMP_COUNT and metrics.SmuMetrics.AvgTemperature[k] != 0]
-    if compact: temps_keys = [(k, name) for k, name in temps_keys if k in (dev.smu.smu_mod.TEMP_HOTSPOT, dev.smu.smu_mod.TEMP_MEM)]
-    return {name: metrics.SmuMetrics.AvgTemperature[k] for k, name in temps_keys}
-
-  def get_voltage(self, dev, metrics, compact=False):
-    voltage_keys = [(k, name) for k, name in dev.smu.smu_mod.c__EA_SVI_PLANE_e__enumvalues.items()
-                        if k < dev.smu.smu_mod.SVI_PLANE_COUNT and metrics.SmuMetrics.AvgVoltage[k] != 0]
-    return {name: metrics.SmuMetrics.AvgVoltage[k] for k, name in voltage_keys}
-
-  def get_busy_threshold(self, dev):
-    match dev.ip_ver[am.MP1_HWIP]:
-      case (14, 0, 2): return 5
-      case _: return 15
-
-  def get_gfx_freq(self, dev, metrics):
-    return metrics.SmuMetrics.AverageGfxclkFrequencyPostDs if self.get_gfx_activity(dev, metrics) <= self.get_busy_threshold(dev) else \
-          metrics.SmuMetrics.AverageGfxclkFrequencyPreDs
-
-  def get_mem_freq(self, dev, metrics):
-    return metrics.SmuMetrics.AverageMemclkFrequencyPostDs if self.get_mem_activity(dev, metrics) <= self.get_busy_threshold(dev) else \
-           metrics.SmuMetrics.AverageMemclkFrequencyPreDs
-
-  def get_fckl_freq(self, dev, metrics):
-    return metrics.SmuMetrics.AverageFclkFrequencyPostDs if self.get_mem_activity(dev, metrics) <= self.get_busy_threshold(dev) else \
-           metrics.SmuMetrics.AverageFclkFrequencyPreDs
-
-  def get_fan_rpm_pwm(self, dev, metrics): return metrics.SmuMetrics.AvgFanRpm, metrics.SmuMetrics.AvgFanPwm
-
-  def get_power(self, dev, metrics): return metrics.SmuMetrics.AverageSocketPower, metrics.SmuMetrics.dGPU_W_MAX
-
  def draw(self):
    terminal_width, terminal_height = shutil.get_terminal_size()
    if self.prev_terminal_width != terminal_width or self.prev_terminal_height != terminal_height:
      os.system('clear')
    self.prev_terminal_width, self.prev_terminal_height = terminal_width, terminal_height

-    padding = 8
-    col_size = (terminal_width) // 2 - padding
-    activity_line_width = 50 if terminal_width > 170 else \
-                         (30 if terminal_width > 130 else \
-                         (16 if terminal_width > 92 else \
-                         max(0, terminal_width - 77)))
+    activity_line_width = 50 if self.prev_terminal_width > 170 else \
+                         (30 if self.prev_terminal_width > 130 else \
+                         (16 if self.prev_terminal_width > 92 else \
+                         max(0, self.prev_terminal_width - 77)))
+    max_col_size = terminal_width // 2

    dev_metrics = self.collect()
    dev_content = []
    for dev, metrics in dev_metrics.items():
      if dev.pci_state != "D0":
        dev_content.append([f"{colored('(sleep)', 'yellow')} {bold(dev.pcibus)}: {self.lspci[dev.pcibus[5:]]}"] +
-                           [pad(f"PCI State: {dev.pci_state}", col_size)])
+                           [f"PCI State: {dev.pci_state}"] + [" "*107])
        continue

-      device_line = [f"{bold(dev.pcibus)} {trim(self.lspci[dev.pcibus[5:]], col_size - 20)}"] + [pad("", col_size)]
-      activity_line = [f"GFX Activity {draw_bar(self.get_gfx_activity(dev, metrics) / 100, activity_line_width)}"] \
-                    + [f"MEM Activity {draw_bar(self.get_mem_activity(dev, metrics) / 100, activity_line_width)}"]
+      device_line = [f"{bold(dev.pcibus)} {trim(self.lspci[dev.pcibus[5:]], max_col_size - 24)}"] + [""]
+      activity_line = [f"GFX Activity {draw_bar(metrics.SmuMetrics.AverageGfxActivity / 100, activity_line_width)}"] \
+                    + [f"MEM Activity {draw_bar(metrics.SmuMetrics.AverageUclkActivity / 100, activity_line_width)}"]

-      temps_data, temps_data_compact = self.get_temps(dev, metrics), self.get_temps(dev, metrics, compact=True)
-      temps_table = ["=== Temps (°C) ==="] + [f"{name:<16}: {color_temp(val)}" for name, val in temps_data.items()]
-      temps_table_compact = ["Temps (°C):" + '/'.join([f"{color_temp(val)} {name}" for name, val in temps_data_compact.items()])]
+      # draw_metrics_table(metrics, dev)
+      temps_keys = [(k, name) for k, name in smu_v13_0_0.c__EA_TEMP_e__enumvalues.items()
+                                if k < smu_v13_0_0.TEMP_COUNT and metrics.SmuMetrics.AvgTemperature[k] != 0]
+      temps_table = ["=== Temps (°C) ==="] + [f"{name:<15}: {color_temp(metrics.SmuMetrics.AvgTemperature[k])}" for k, name in temps_keys]
+      temps_table_compact = [f"Temps (°C): {color_temp(metrics.SmuMetrics.AvgTemperature[smu_v13_0_0.TEMP_HOTSPOT])} hotspot /" \
+                          + f"{color_temp(metrics.SmuMetrics.AvgTemperature[smu_v13_0_0.TEMP_MEM])} mem"]

-      fan_rpm, fan_pwm = self.get_fan_rpm_pwm(dev, metrics)
-      power_table = ["=== Power ==="] + [f"Fan Speed: {fan_rpm} RPM"] + [f"Fan Power: {fan_pwm}%"]
+      voltage_keys = [(k, name) for k, name in smu_v13_0_0.c__EA_SVI_PLANE_e__enumvalues.items() if k < smu_v13_0_0.SVI_PLANE_COUNT]
+      power_table = ["=== Power ==="] \
+                  + [f"Fan Speed: {metrics.SmuMetrics.AvgFanRpm} RPM"] \
+                  + [f"Fan Power: {metrics.SmuMetrics.AvgFanPwm}%"]
+      power_line = [f"Power: {metrics.SmuMetrics.AverageSocketPower:>3}W " +
+                       draw_bar(metrics.SmuMetrics.AverageSocketPower / metrics.SmuMetrics.dGPU_W_MAX, 16)]
+      power_line_compact = [f"Power:  {metrics.SmuMetrics.AverageSocketPower:>3}W " +
+                       draw_bar(metrics.SmuMetrics.AverageSocketPower / metrics.SmuMetrics.dGPU_W_MAX, activity_line_width)]

-      total_power, max_power = self.get_power(dev, metrics)
-      power_line = [f"Power: {total_power:>3}W " + draw_bar(total_power / max_power, 16)]
-      power_line_compact = [f"Power:  {total_power:>3}W " + draw_bar(total_power / max_power, activity_line_width)]
+      voltage_table = ["=== Voltages ==="] + [f"{name:<20}: {color_voltage(metrics.SmuMetrics.AvgVoltage[k])}" for k, name in voltage_keys]

-      voltage_data = self.get_voltage(dev, metrics)
-      voltage_table = ["=== Voltages ==="] + [f"{name:<20}: {color_voltage(voltage)}" for name, voltage in voltage_data.items()]
+      gfx_freq = (metrics.SmuMetrics.AverageGfxclkFrequencyPostDs if metrics.SmuMetrics.AverageGfxActivity <= SMU_11_0_GFX_BUSY_THRESHOLD else \
+                  metrics.SmuMetrics.AverageGfxclkFrequencyPreDs)
+      fclk_freq = (metrics.SmuMetrics.AverageFclkFrequencyPostDs if metrics.SmuMetrics.AverageUclkActivity <= SMU_11_0_GFX_BUSY_THRESHOLD else \
+                   metrics.SmuMetrics.AverageFclkFrequencyPreDs)
+      mclk_freq = (metrics.SmuMetrics.AverageMemclkFrequencyPostDs if metrics.SmuMetrics.AverageUclkActivity <= SMU_11_0_GFX_BUSY_THRESHOLD else \
+                   metrics.SmuMetrics.AverageMemclkFrequencyPreDs)

-      gfx_freq = self.get_gfx_freq(dev, metrics)
-      mclk_freq = self.get_mem_freq(dev, metrics)
-      fclk_freq = self.get_fckl_freq(dev, metrics)
+      frequency_table = ["=== Frequencies ===",
+        f"GFXCLK Target : {metrics.SmuMetrics.AverageGfxclkFrequencyTarget:>4} MHz",
+        f"GFXCLK PreDs  : {metrics.SmuMetrics.AverageGfxclkFrequencyPreDs:>4} MHz",
+        f"GFXCLK PostDs : {metrics.SmuMetrics.AverageGfxclkFrequencyPostDs:>4} MHz",
+        f"FCLK PreDs    : {metrics.SmuMetrics.AverageFclkFrequencyPreDs:>4} MHz",
+        f"FCLK PostDs   : {metrics.SmuMetrics.AverageFclkFrequencyPostDs:>4} MHz",
+        f"MCLK PreDs    : {metrics.SmuMetrics.AverageMemclkFrequencyPreDs:>4} MHz",
+        f"MCLK PostDs   : {metrics.SmuMetrics.AverageMemclkFrequencyPostDs:>4} MHz",
+        f"VCLK0         : {metrics.SmuMetrics.AverageVclk0Frequency:>4} MHz",
+        f"DCLK0         : {metrics.SmuMetrics.AverageDclk0Frequency:>4} MHz",
+        f"VCLK1         : {metrics.SmuMetrics.AverageVclk1Frequency:>4} MHz",
+        f"DCLK1         : {metrics.SmuMetrics.AverageDclk1Frequency:>4} MHz"]

-      frequency_table = ["=== Frequencies ===", f"GFXCLK: {gfx_freq:>4} MHz", f"FCLK  : {fclk_freq:>4} MHz", f"MCLK  : {mclk_freq:>4} MHz"]
+      frequency_table_compact = ["=== Frequencies ===",
+        f"GFXCLK: {gfx_freq:>4} MHz",
+        f"FCLK  : {fclk_freq:>4} MHz",
+        f"MCLK  : {mclk_freq:>4} MHz"]

      if self.prev_terminal_width >= 231:
        power_table += power_line + [""] + voltage_table
        activity_line += [""]
      elif self.prev_terminal_width >= 171:
-        power_table += power_line + [""] + frequency_table
+        power_table += power_line + [""] + frequency_table_compact
        activity_line += [""]
        frequency_table = None
      elif self.prev_terminal_width >= 121:
        temps_table = None
+        frequency_table = frequency_table_compact
        activity_line += power_line_compact
      else:
        temps_table = None
@@ -240,7 +222,7 @@ class SMICtx:

    raw_text = 'AM Monitor'.center(terminal_width) + "\n" + "=" * terminal_width + "\n\n"
    for i in range(0, len(dev_content), 2):
-      if i + 1 < len(dev_content): raw_text += '\n'.join(same_line([dev_content[i], dev_content[i+1]], split=padding))
+      if i + 1 < len(dev_content): raw_text += '\n'.join(same_line([dev_content[i], dev_content[i+1]]))
      else: raw_text += '\n'.join(dev_content[i])
      if i + 2 < len(dev_content): raw_text += "\n" + "=" * terminal_width + "\n\n"

--- a/Show More
+++ b/Show More