Commit Graph

Select branches

Hide Pull Requests

abstractions_4

add_local_buffer

after_qol

algebraic_upat

all_bv_no_idx

all_sink

amd_isel

amd_refac

amd_smu_reset_patch

amd_sqtt

amd_uop

amd_uop_cleanups

assign_copy

assign_try_3

autogen-comgr-2

before2

better_spec

better_syn

bitcast_spec

broadcast_shape_2

bt_ei

buf_heur

bump_nv_2404

cache_reshape

call_inline

call_is_kernels

callify

cast_vec_as_cast

cat_mop

changes_from_rangify

check_clsetarg

check_dims

ci-deps

cifar_30

cifar_30_clean

clean_load

clean_spec

clone_tg

codegen_try_2

contig_on_after

continue_mat

deepsek

default_late_allreduce

delete_kernel_py

delete_ones

delete_slow_rangeify

demote_op

disable_rules

dsp_search

dsp_search_merged

dtype_shape

earlier_gater

emu_cdna

emulate_ctx

end_all_ranges

endrange

er_prioity

execution_unit

explicit-dev

failing_assign

fancy_usb

fast_llm

fix_fa_fwd_clang_22

fix_fuse

fix_llvm_cast

fix_range_merging_sd

fix_rangeify_tests

fix_rc

fix_sched_copy

fix_slice_store

fix_src_spec

fixup_fix_kernel

full_spec

fun_w_egraphs

gate-on-load-store

gate_rewrite

gemma4_gpt

gen_pdf_fast

gh-pages

gitea-test-ci

gl_dims

gldims

glm_flash

grad_outerworld

grok_x86_novec

group_div_rules

heur_4096

heur_group

hk_gemm

http_threading

hwtest_fixes

id_store

if_range_cleanups

ignore_oob

image_idx_xy

image_in_mc

image_no_vec

image_refactors_stack

index_to_slice

invalid_try_3

jinja_template

kernel4

kernel_is_call

kernelless

kimi_slop

kitten_matmul

late_Locals

late_add_load

late_allreduce

late_codegen

late_decomp_is_late

late_lower_idx

late_view

late_weakint

lds_double_buffering

less_wmma_arg

lil_dg_change

lil_rangeify_td

llama_trainer

llm_262k_minimal

llm_flash_attn

llm_qwen

llm_serve_fixes

load_view_late

local_cache

locals_have_device

loop_end_op

loop_is_rng

lowerer_hacks

mac_pytest

manual_kernel_3

master

mega_lowerer

min_new_lin

min_outer_range

minor_views

mixins7

mlperf_training_v4.1

mlperf_training_v5.0

mlperf_training_v6.0

modern_scan

more_apps

more_early_comps

more_folding_rules

more_kernelcount

more_pcode

more_rangeify_pm_tests

more_ren_clean

more_uop_gc_test

more_uop_mm

more_vmap

move_devec

move_dtype_decomps

move_expander

move_gates_to_load_store

move_mop_cleanup

move_schedule

move_simp_views

moveleftright

mt_typing

multi_range

multioutput

new_ck

new_devec

new_invalid_and_image

new_linearizer

new_shape

new_test

no-chat-template

no_assign_2

no_buf_0

no_decimals

no_late_simplify_marg

no_merge_ends

no_merge_views

no_mstack_in_allreduce

no_one_ranges

no_sip_usbgpu

no_tsink_base

no_tuplize

no_user_ck_contig

no_vec_x86_4

noop_index

novecimg

num_batches_tracked_shape

olmo3_support

one_compiler

one_hcopt

only_assign_buffers

only_loop_backends

only_reg_emu

only_reg_emu2

onnx-cpu-count

outer_range

outerworld_work

pad_invalid

pad_invalid_work

pad_mixin

param_call_is_function

parse_sqtt_more

pci_iommu

port_nir

postrange_hcopts

precompile_backward

prefill_qwen_fast

print_special_ops

ps_clean

pyremu_try

python_prg

python_speed

q4k_support

qcom-openpilot-13ms

qcom_830

qcom_mmapeak

qcom_speed_codex

qol_string

qwen36_27b_amd_900

qwen36_27b_amd_minimal

qwen36_amd_bugfixes

r_cleanups

r_new_opt

ranged_call

rangeify_2

rangeify_const_folding

rdna3_vibes

rdna4_asm

rdna4_gemm

real_pyrender

recurse_sub

reg_coalese

relu_pattern

remove_assign

remove_cast_hack

remove_ck_contig

remove_define_var

remove_forced_re

remove_on_stack

remove_programspec

remove_pylint

remove_slow_tests

remove_vec_2

remove_vec_3

renderers_use_shapes_2

replace_if_with_range

reprocess_node_experiments

reshape_trait

resnet-ci

revert-11547-fix_mismatch_red

revert-11715-revert-11711-amd_llvm_tc_cast

revert-11809-no_ki_in_gpudims

revert-11811-better_div_nesting

revert-11876-memplan_vbig

revert-12104-simplify_valid_in_symbolic

revert-12308-bufferize_early

revert-12731-jitlink_init_check

revert-12813-after_cleanups

revert-13046-better_cleanup_arange_buffers

revert-13311-prof_link_sqtt

revert-13583-rocr_desc_2

revert-14734-tload

revert-17141-full_rewrite_to_sink

revert-17408-_broadcasted-weak

revert-9438-simple_matmul_np_rand

revert_wrong_ck_fix

rewrite_rangeify2

rtoposort

scan_assign

scan_op_8

sched_cache

sched_lin

sd_load_simple

self_type

set_testpath

shrink_in_render

simp_spec

simple_gidxs

simple_pool

simple_priority

simple_rangeify

simpler-sd-compile

simpler_end

simpler_fusion

simpler_image

simpler_postrange

simpler_reduce

slice_like_shrink

spec_3

spec_cleanup

spec_ren

speed_one_linear

split_buf_idx

split_decomp_pass

split_sub

stack_0

store_is_void

string_ranges

support_opts_in_contig

swizzle_in_kernel

sym_llm

sym_work

symbolic-chunk-split

szdiff_bot_test

tc_is_heur

tensor_graph_single_sink

test-lp

test_amd_crash

test_fa

test_from_postopt

test_hgemm

test_unit

tiny_scan

tmr

tool_call_support

topoprop

ttomsa-new_linearizer

tuplegtuple

two_stage_remove

typed_checks

typing_work

unused_permute_arg

uop_prg

upd_test4

update_benchmark

update_benchmark_search

update_mlperf

user_failure

vecless_x86_glm

verify_sink_dims_clean

view_after_const

view_in_codegen

viz_cbroadcast

warp_fun

wgpu-f16-packed

working_postopt

x86_moves

x86_numel

#1

#2

#3

#4

#5

v0.10.0

v0.10.1

v0.10.2

v0.10.3

v0.11.0

v0.12.0

v0.13.0

v0.4.0

v0.5.0

v0.6.0

v0.7.0

v0.8.0

v0.9.0

v0.9.1

v0.9.2

Mono Color

• d462f8ace0 use HALF in cifar wino benchmarks (#8153) chenyuandGitHub 2024-12-10 20:21:00 -05:00
• c8e7707a7e hotfix: disable flaky move tensor test geohot 2024-12-10 17:11:21 -08:00
• 155f7df599 lower test_gemm_4096 expectation on green (#8152) chenyuandGitHub 2024-12-10 18:05:12 -05:00
• c4be1529cf update test for Tensor.softplus (#8150) chenyuandGitHub 2024-12-10 17:48:02 -05:00
• a8cfdc70ed Run more webgpu tests (#8142) Ahmed HarmoucheandGitHub 2024-12-10 23:20:04 +01:00
• ed7318a3f5 Fix puppeteer install (#8148) Ahmed HarmoucheandGitHub 2024-12-10 23:06:33 +01:00
• a1b3724ff8 prepickle process replay [pr] (#8147) George HotzandGitHub 2024-12-10 11:46:36 -08:00
• aa3b094334 changes from delete lazy [pr] (#8146) George HotzandGitHub 2024-12-10 11:06:17 -08:00
• 286fec115e fix Tensor.minimum for int (#8145) chenyuandGitHub 2024-12-10 13:34:41 -05:00
• 71dd222f66 Fix setitem on wgpu (#8144) Ahmed HarmoucheandGitHub 2024-12-10 19:34:25 +01:00
• b69fea6ae5 process replay without global list [pr] (#8143) qazalandGitHub 2024-12-10 20:20:09 +02:00
• 08405279f9 pre merge_views+ops_folding refactor [pr] (#8140) qazalandGitHub 2024-12-10 18:55:00 +02:00
• 56c84cee29 derive COPY nbytes late in realize [pr] (#8137) qazalandGitHub 2024-12-10 16:04:07 +02:00
• 2d26b011ac allow VIEW on BUFFER [pr] (#8136) qazalandGitHub 2024-12-10 15:29:38 +02:00
• 3a2658efbd small changes to refine the delete_lazy diff (#8134) qazalandGitHub 2024-12-10 12:46:10 +02:00
• 6d33da09c9 split scalar getitem tests into correctness and optimization [pr] (#8133) qazalandGitHub 2024-12-10 12:18:46 +02:00
• 7436ebef2f spend lines on const_arg for tensor and scheduler [pr] (#8132) qazalandGitHub 2024-12-10 12:07:35 +02:00
• 917deb88a4 make //0 return 0 in python_alu (#8131) chenyuandGitHub 2024-12-09 19:32:06 -05:00
• f83d715f41 move checks into compile3, delete compile2 [pr] (#8127) George HotzandGitHub 2024-12-10 06:21:42 +08:00
• 358287959b fix pow of int to negative const int (#8129) chenyuandGitHub 2024-12-09 17:20:18 -05:00
• 12f7d284e0 failed test case for int pow (#8128) chenyuandGitHub 2024-12-09 16:15:09 -05:00
• 80de06c8b9 scheduler ops_folding from delete_lazy (#8124) qazalandGitHub 2024-12-09 18:36:01 +02:00
• 87c360c4b5 hotfix: add --size 8B to llama3 geohot 2024-12-09 07:53:20 -08:00
• a773c5a571 hotfix: default llama3 is 1B with download_model geohot 2024-12-09 07:23:35 -08:00
• c6277fce09 Remove f16 decompression lib from SD compile.py (#8121) Ahmed HarmoucheandGitHub 2024-12-09 14:09:00 +01:00
• 22d99f1421 test_pickle_realized_tensor actually tests pickle [pr] (#8119) qazalandGitHub 2024-12-09 11:26:19 +02:00
• ccf54c2375 fix argmax/min on int32 min (#8118) chenyuandGitHub 2024-12-09 02:29:23 -05:00
• c814de2dd4 fix bitwise_not for signed int (#8117) chenyuandGitHub 2024-12-09 02:02:51 -05:00
• e22d7b6fb0 fix var vmax inside special (#8116) ttomsaandGitHub 2024-12-09 06:16:08 +00:00
• 0033012096 init noop changes from delete_lazy [pr] (#8115) qazalandGitHub 2024-12-08 19:42:05 +02:00
• 5dd61035f7 revert VALID early folding for now (#8114) qazalandGitHub 2024-12-08 18:34:24 +02:00
• 69e48da961 set NOOPT in test_avg_pool3d_failure (#8112) qazalandGitHub 2024-12-08 17:48:29 +02:00
• 3a7d64b96c hcq remove update from args state (#8104) nimlgenandGitHub 2024-12-08 15:22:05 +03:00
• d6e66095fd hcq buffer is a class (#8106) nimlgenandGitHub 2024-12-08 13:29:43 +03:00
• b9c977f1c8 clean up bounds in Tensor.shard (#8107) chenyuandGitHub 2024-12-07 17:19:43 -05:00
• f8294b3bda add avg pool 3d failure test (#8105) geohotstanandGitHub 2024-12-08 05:34:38 +08:00
• 6be388be86 failing test for const folding breaking indexing [pr] (#8103) qazalandGitHub 2024-12-07 13:55:02 +02:00
• 8b1fa9cb7d nv hcq queue touchups (#8102) nimlgenandGitHub 2024-12-07 14:09:38 +03:00
• 4074f52317 VALID early folding (#8100) qazalandGitHub 2024-12-07 12:37:47 +02:00
• 07b6d5cf63 assign early folding (#8093) qazalandGitHub 2024-12-07 11:02:55 +02:00
• 00ac0db9d4 np tensors have the memory from numpy in compile3 [pr] (#8098) George HotzandGitHub 2024-12-07 14:01:51 +08:00
• 22feb3a2f1 move copy into the JIT for openpilot compile3 (#7937) George HotzandGitHub 2024-12-07 13:26:26 +08:00
• 0ed731b5ea torch_load with Tensors (#8037) leopfandGitHub 2024-12-07 02:55:41 +01:00
• 2d321646b8 default tensors to int32 in test_ops (#8097) chenyuandGitHub 2024-12-06 20:33:36 -05:00
• e9692de42b don't FUZZ_ALL_ACTIONS in fuzz_linearizer.py (#8096) chenyuandGitHub 2024-12-06 17:22:17 -05:00
• 564b3a3e1b onnx Bitwise ops (#8095) chenyuandGitHub 2024-12-06 16:58:09 -05:00
• a97b8fa3c5 maskless const can lower without valid, p1 [pr] (#8094) qazalandGitHub 2024-12-06 23:21:19 +02:00
• aaf2379f97 remove ordered parents, seems like dead code [pr] (#8092) mesozoic-eggandGitHub 2024-12-07 05:19:37 +08:00
• e180a31c5e tiny metal cleanup (#8089) nimlgenandGitHub 2024-12-06 21:44:32 +03:00
• d000c08f04 fix return type of Tensor.pow (#8091) chenyuandGitHub 2024-12-06 13:38:29 -05:00
• 1ea4dc9565 big graph init conceptual cleanup [pr] (#8090) qazalandGitHub 2024-12-06 20:07:00 +02:00
• 5184410fc3 combine get inputs and type_parse function in onnx [fixed] (#8081) 3 people 2024-12-07 01:34:47 +08:00
• d1282da7e8 hcq bump alloc (#8078) nimlgenandGitHub 2024-12-06 19:19:04 +03:00
• df84dc6444 unrelated test fixups from delete_lazy [pr] (#8088) qazalandGitHub 2024-12-06 17:31:02 +02:00
• 0b7c44677d Fix uint8 cast underflow (#6305) 3 people 2024-12-06 23:25:03 +08:00
• f3983f6743 Move efficientnet example (#8087) Ahmed HarmoucheandGitHub 2024-12-06 15:48:16 +01:00
• 7dbd166227 skip test_schedule_mem_used_with_inputs [pr] (#8086) qazalandGitHub 2024-12-06 16:44:34 +02:00
• 0356657ced move view_supported_devices to device [pr] (#8085) qazalandGitHub 2024-12-06 16:44:15 +02:00
• fad3eaa35e Use atomicLoad builtin when loading atomic type (#8084) Ahmed HarmoucheandGitHub 2024-12-06 15:33:11 +01:00
• 79966fade0 free up lines for const_arg [pr] (#8083) qazalandGitHub 2024-12-06 16:28:51 +02:00
• ba35c4138b Use matching JS TypedArray for buffer dtype (#8080) Ahmed HarmoucheandGitHub 2024-12-06 14:52:23 +01:00
• a684d72e55 add ceil_mode for avg_pool and max_pool (#7579) 3 people 2024-12-06 21:34:14 +08:00
• b73d9a7d24 Revert "combine get inputs and type_parse function in onnx (#8069)" (#8079) chenyuandGitHub 2024-12-06 08:04:21 -05:00
• c8313a3669 Cleaner rule for mul/idiv by power of two [pr] (#8076) Sieds LyklesandGitHub 2024-12-06 14:02:24 +01:00
• a77ee72d11 clean up reshape size check [pr] (#8067) chenyuandGitHub 2024-12-06 07:51:19 -05:00
• 074a67a6eb combine get inputs and type_parse function in onnx (#8069) geohotstanandGitHub 2024-12-06 20:42:35 +08:00
• c0240855b9 qcom has not transfer (#8075) nimlgenandGitHub 2024-12-06 14:45:01 +03:00
• ce72fe1411 u32 to f16 in tinygrad (#8074) Ahmed HarmoucheandGitHub 2024-12-06 12:00:13 +01:00
• e37bff6c19 fix bug in jit prune with copy [pr] (#8073) George HotzandGitHub 2024-12-06 18:38:23 +08:00
• aae8557ada test copy inside jit [pr] (#8072) George HotzandGitHub 2024-12-06 17:51:50 +08:00
• e2fe7f0d2f hotfix: actually fix pylint, it's a python 3.10 issue geohot 2024-12-06 13:53:46 +08:00
• b28d660172 update self_tokenize, fix pylint maybe geohot 2024-12-06 13:49:41 +08:00
• 344fd4845c example: self_tokenize. someday tinygrad will be recursively self improving geohot 2024-12-06 13:35:02 +08:00
• 3c5d5f9414 mypy==1.13.0 (#7990) 3 people 2024-12-06 05:09:14 +01:00
• 65b6696f3b refactor safe_load (#8035) leopfandGitHub 2024-12-06 05:08:21 +01:00
• e7d5fe4a32 improve idiv _min_max (#8066) chenyuandGitHub 2024-12-05 23:02:16 -05:00
• 13b954f22c unify expand conditions [pr] (#8065) chenyuandGitHub 2024-12-05 21:40:14 -05:00
• aefdff4ef5 reshape mask cleanups [pr] (#8064) chenyuandGitHub 2024-12-05 20:20:43 -05:00
• 05dba6e4ee minor to_indexed_uops cleanup [pr] (#8063) chenyuandGitHub 2024-12-05 17:15:03 -05:00
• b2dd703592 fix typing of UOp.range [pr] (#8062) chenyuandGitHub 2024-12-05 14:56:34 -05:00
• 49c6dab74b Add pattern for div mod recombine with gcd (#8061) 3 people 2024-12-05 19:16:58 +01:00
• 707e9a9c8e add _one_hot_along_dim helper for Tensor.arange masking (#8039) 3 people 2024-12-06 01:43:00 +08:00
• 3c5983473a combine parentless reduce rule [pr] (#8059) chenyuandGitHub 2024-12-05 11:28:35 -05:00
• 87594a8153 simpler dtypes.max for int [pr] (#8058) chenyuandGitHub 2024-12-05 10:31:41 -05:00
• 66b8242375 Simple onnx.py clean ups (#8054) 3 people 2024-12-05 23:31:26 +08:00
• 5c6ed5dba6 lower test_conv_3x3_256_32_32_256_256 expectation (#8060) chenyuandGitHub 2024-12-05 10:30:56 -05:00
• c6f5bb03fa YoloV8 WebGPU fixes (#8057) Ahmed HarmoucheandGitHub 2024-12-05 16:23:45 +01:00
• fc9ae53a05 Progress with pack/unpack wgsl built-ins wgpu-f16-packed Ahmed Harmouche 2024-12-05 15:55:09 +01:00
• 4669e467ef f16 comp/decomp WIP Ahmed Harmouche 2024-12-05 14:14:23 +01:00
• 78c01a5c2b amd general _gpu_alloc (#8056) nimlgenandGitHub 2024-12-05 15:50:23 +03:00
• 8071600897 nv one _gpu_alloc (#8055) nimlgenandGitHub 2024-12-05 15:22:03 +03:00
• ff9a89f714 Proper dtypes for input/output of exported WebGPU model (#8053) Ahmed HarmoucheandGitHub 2024-12-05 10:38:05 +01:00
• 435a51e10c reduce folding simple tests [pr] (#8040) qazalandGitHub 2024-12-04 23:22:45 -05:00
• 20878be2af lower test_gemv_4096_16384 expectations geohot 2024-12-05 12:07:38 +08:00
• 83aecbdc70 do gpuocelot copy manually [pr] (#8050) George HotzandGitHub 2024-12-05 11:51:20 +08:00
• 4a208bfb28 bump download cache version geohot 2024-12-05 11:42:34 +08:00
• df18e7cc37 accept filename decorator [pr] (#8049) George HotzandGitHub 2024-12-05 11:40:59 +08:00
• c3187087f7 QwQ-32B-Preview support (#7962) Francis LataandGitHub 2024-12-04 21:46:37 -05:00
• b3220ca7b1 test cases of always True/False lt (#8048) chenyuandGitHub 2024-12-04 20:38:40 -05:00
• 8bb806888b hook_overflow -> safe_exp2 [pr] (#8047) chenyuandGitHub 2024-12-04 19:05:38 -05:00