geohot/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests Actions Activity

cuda fp8 (#12782)

Browse Source 
* cuda fp8

* tensor core

* tc test

* clean

* clean pm
This commit is contained in:
b1tg
2025-10-21 15:05:25 -04:00
committed by GitHub
parent 587ccc0e5c
commit 60d7e232f2
7 changed files with 43 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/test.yml
+1 -1
View File
@@ -204,7 +204,7 @@ jobs:
DEBUG=2 EMULATE=CUDA FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm_fp16
DEBUG=2 EMULATE=CUDA ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
DEBUG=2 EMULATE=CUDA_SM75 FORWARD_ONLY=1 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm_fp16
DEBUG=2 EMULATE=CUDA ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/opt/test_tensor_cores.py
DEBUG=2 EMULATE=CUDA_SM89 ALLOW_TF32=1 FORWARD_ONLY=1 PYTHON=1 python3 test/opt/test_tensor_cores.py
- name: Test emulated INTEL OpenCL tensor cores
run: DEBUG=2 EMULATE=INTEL FORWARD_ONLY=1 PYTHON=1 HALF=1 N=64 python3 ./extra/gemm/simple_matmul.py
- name: Test emulated AMX tensor cores