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

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

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

delta_attention

delta_attention_scan

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

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

gl_dims

gldims

glm_flash

grad_outerworld

grok_x86_novec

group_div_rules

hcq2_vram_copies

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_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

no_weakint_variable

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

outer_range

outerworld_work

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

qwen_mergable

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_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

vf_ish

view_after_const

view_in_codegen

viz_cbroadcast

warp_fun

wgpu-f16-packed

working_postopt

x86_moves

x86_numel

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

• 9119716761 update Tensor.maximum (#8992) chenyuandGitHub 2025-02-09 21:26:27 -05:00
• 88add71c25 amd: increase sdma copy size (#8989) nimlgenandGitHub 2025-02-09 20:53:35 +03:00
• 7eba5fb413 Tensor.empty is RESHAPE(BUFFER) (#8987) qazalandGitHub 2025-02-09 18:42:51 +01:00
• 44479f8ad6 raise ValueError in view reshape for negative dims [pr] (#8988) qazalandGitHub 2025-02-09 17:27:15 +01:00
• c6c2373bc0 replace libpciaccess autogen with just pci regs (#8983) nimlgenandGitHub 2025-02-09 18:40:45 +03:00
• 55351ebb31 minimal failing test for #8975 [pr] (#8982) qazalandGitHub 2025-02-09 14:10:37 +01:00
• e5a3f60fc2 am: remove libpciaccess dep (#8980) nimlgenandGitHub 2025-02-09 16:06:55 +03:00
• 52a69dd5e9 Revert "use am in training benchmarks (#8965)" (#8981) nimlgenandGitHub 2025-02-09 15:43:45 +03:00
• 0b26cee2f1 fix some slow tests [pr] (#8979) George HotzandGitHub 2025-02-09 15:57:04 +08:00
• 208097d488 try reducing testing deps [pr] (#8976) George HotzandGitHub 2025-02-09 15:22:32 +08:00
• 6ffee2fca9 reduce speed example [pr] (#8978) George HotzandGitHub 2025-02-09 14:13:59 +08:00
• ac3765c043 use getpass instead of os.getlogin() (#8972) Samuel AyalaandGitHub 2025-02-08 15:29:26 -05:00
• 308516e439 fix viz paginate + cleanups [pr] (#8973) qazalandGitHub 2025-02-08 20:26:57 +01:00
• 107e616857 use am in training benchmarks (#8965) nimlgenandGitHub 2025-02-08 20:20:47 +03:00
• 79de980565 am: do not fork pci bars (#8969) nimlgenandGitHub 2025-02-08 19:03:17 +03:00
• 0cac941af1 move xpow to sym instead of late_rewrite (#8968) chenyuandGitHub 2025-02-08 10:09:24 -05:00
• e7182bbb2c fix "fatal bad object" log in process replay [pr] (#8966) qazalandGitHub 2025-02-08 11:57:38 +01:00
• 9b9c1e14da Late MTLCompiler load (#8963) uuuvnandGitHub 2025-02-08 14:29:23 +05:00
• a3c78d47b3 speed docs + upgrades [pr] (#8964) George HotzandGitHub 2025-02-08 17:28:52 +08:00
• 5bdd6a1cc4 increase CI speed with more runners [pr] (#8961) George HotzandGitHub 2025-02-08 09:04:36 +08:00
• 11d50324d8 am: tiny cleanups (#8958) nimlgenandGitHub 2025-02-07 23:44:43 +03:00
• cfd28517df move pow folding tests to test_schedule [pr] (#8955) chenyuandGitHub 2025-02-07 12:51:43 -05:00
• c2b4c43edb handle stride 0 reduce (#8068) 3 people 2025-02-07 22:40:58 +08:00
• cf21e27d78 little better VIEW simplifier pattern [pr] (#8954) qazalandGitHub 2025-02-07 12:55:54 +01:00
• 329013f577 fix UOp.metadata on KERNEL op [pr] (#8953) qazalandGitHub 2025-02-07 12:40:11 +01:00
• 4de084a835 cleanup ci, split docs/autogen, testing_minimal, LLVM Speed [pr] (#8952) George HotzandGitHub 2025-02-07 19:01:59 +08:00
• 6090cbe3be Try to open llvm first when opening metal (#8949) uuuvnandGitHub 2025-02-07 15:58:37 +05:00
• 67b70e4f6c Fix incorrect __del__ (#8950) uuuvnandGitHub 2025-02-07 15:21:44 +05:00
• 9ed2d0dfa2 refactor into subactions (#8946) George HotzandGitHub 2025-02-07 18:06:44 +08:00
• 133cacadde Autogen webgpu dawn, removing wgpu-py dependency (f16 support part 1) (#8646) 3 people 2025-02-07 08:16:59 +01:00
• dbda72f91d hotfix: raise line limit to 11200 for new webgpu backend geohot 2025-02-07 14:29:20 +08:00
• b1e1319972 ci speed on the enterprise plan [pr] (#8942) George HotzandGitHub 2025-02-07 11:18:12 +08:00
• 3b67712892 [bounty] Fix LLVM=1 NO_DEVECTORIZE=1 python3 test/test_ops.py TestOps.test_strided_conv2d_simple (#8937) 3 people 2025-02-06 18:07:54 -08:00
• f54242849d failing test for the devectorize [pr] (#8940) George HotzandGitHub 2025-02-07 09:44:54 +08:00
• ee1a0fb8ec am_smi: print device name (#8939) nimlgenandGitHub 2025-02-07 03:01:25 +03:00
• a092b6395d Tuple -> tuple, List -> list [pr] (#8936) chenyuandGitHub 2025-02-06 14:21:19 -05:00
• d5183e1584 remove unneeded annotation import (#8934) chenyuandGitHub 2025-02-06 13:12:35 -05:00
• 00d72a5144 setitem isinstance cleanup [pr] (#8932) chenyuandGitHub 2025-02-06 11:44:57 -05:00
• 81e241150a hotfix: save 1 line (#8931) qazalandGitHub 2025-02-06 16:26:05 +01:00
• eb1144be8b hotfix: only check current graph when excluding nodes in viz (#8930) qazalandGitHub 2025-02-06 15:58:53 +01:00
• 3cc05081f4 llvm no devectorize, the right way (#8901) George HotzandGitHub 2025-02-06 22:53:49 +08:00
• 8b16c65bca add compile3 benchmark [pr] (#8929) George HotzandGitHub 2025-02-06 22:49:31 +08:00
• 79fb5c6470 hotfix: test_shard_no_recompile shouldn't rely on schedule order [pr] (#8928) qazalandGitHub 2025-02-06 15:27:59 +01:00
• 1249e8dd3b objc fast msg, try 2 [pr] (#8927) George HotzandGitHub 2025-02-06 19:06:21 +08:00
• 86feb98dcd am: add support for 7600 (#8910) nimlgenandGitHub 2025-02-06 14:04:07 +03:00
• ae45826758 hotfix: GRAPH_ONE_KERNEL + fix timing geohot 2025-02-06 17:52:20 +08:00
• 1c53e8bf27 Revert "objc fast msg (#8922)" (#8926) George HotzandGitHub 2025-02-06 17:50:49 +08:00
• c3f99a727e objc fast msg (#8922) George HotzandGitHub 2025-02-06 17:49:06 +08:00
• a2e7e49fe1 prepickle scheduler process replay [pr] (#8924) qazalandGitHub 2025-02-06 10:16:36 +01:00
• 89d7480b0c hotfix: don't sink views [pr] (#8923) qazalandGitHub 2025-02-06 09:15:12 +01:00
• 0cbb7d7f1e hotfix: metal has known sync issue geohot 2025-02-06 14:29:41 +08:00
• a8e54df363 benchmark single kernel launch (#8921) George HotzandGitHub 2025-02-06 13:35:34 +08:00
• 3e082d4a9d add float4 support to LLVM (#8920) George HotzandGitHub 2025-02-06 12:15:50 +08:00
• b05c536f74 cleanup some llvm stuff [pr] (#8919) George HotzandGitHub 2025-02-06 11:45:03 +08:00
• 44e0eab8fd Fix AttributeError occurring after ValueError in _apply_uop (#8905) 3 people 2025-02-05 21:56:29 -05:00
• 30695da256 remove Tensor._to_const_val (#8917) chenyuandGitHub 2025-02-05 21:44:39 -05:00
• d09b5f801c don't use Tensor new, add to all_tensors after constructions [pr] (#8918) George HotzandGitHub 2025-02-06 10:21:32 +08:00
• 759b3f86bf Pass host CPU features to LLVM target (#8909) FICTURE7andGitHub 2025-02-06 06:19:30 +04:00
• 09ec33a578 Better errors when relocating against undefined symbol (#8902) uuuvnandGitHub 2025-02-06 04:13:44 +02:00
• 488200f16c move more pow const to rewrite (#8916) chenyuandGitHub 2025-02-05 20:30:12 -05:00
• 76671381aa move positive const ** t to a rewrite rule (#8914) chenyuandGitHub 2025-02-05 19:30:12 -05:00
• cad44f5f42 add Half-Precision Accumulation Support for Tensor Cores in NV, CUDA, and PTX (#8680) 3 people 2025-02-05 18:56:37 -03:00
• 17f9b1cef6 am: load fw based on versions (#8913) nimlgenandGitHub 2025-02-06 00:02:09 +03:00
• 189bfa164e enable backward test for pow(neg const ** x) (#8912) chenyuandGitHub 2025-02-05 15:35:21 -05:00
• 9307572fe3 Ops.POW and transcendental (#8911) chenyuandGitHub 2025-02-05 15:15:59 -05:00
• bff7c70eef hcq: better var check (#8908) nimlgenandGitHub 2025-02-05 22:38:59 +03:00
• aec3b8d515 add regression test: test_get_kernel_actions_preserves_actions_state (#8907) Ignacio SicaandGitHub 2025-02-05 16:13:01 -03:00
• e71497aabc move assign ShapeTracker check to pattern matcher [pr] (#8906) qazalandGitHub 2025-02-05 19:47:20 +01:00
• 0f6109ec00 hotfix bug in get_kernel_actions after TC_SEARCH_OVER_SHAPE was introduced (#8904) Ignacio SicaandGitHub 2025-02-05 15:10:05 -03:00
• 15f94ac964 TC_SEARCH_OVER_SHAPE to search multiple TC shapes (#8793) Ignacio SicaandGitHub 2025-02-05 13:03:46 -03:00
• e7edadda54 construct the sched_sink with graph_rewrite [pr] (#8903) qazalandGitHub 2025-02-05 15:16:48 +01:00
• ef7ad3f077 simpler subbuffer construction + copyin is always base (#8900) qazalandGitHub 2025-02-05 09:10:20 +01:00
• 6f0cc2e9c5 rename to KernelContext and move the linearize_sched comment [pr] (#8899) qazalandGitHub 2025-02-05 07:49:58 +01:00
• 6fb0e5751b hotfix test_onnx_imagenet (#8897) geohotstanandGitHub 2025-02-05 14:39:55 +08:00
• c1c5227acb preserve size in dtype ptr [pr] (#8898) George HotzandGitHub 2025-02-05 14:38:57 +08:00
• 5844883e59 bump master version v0.10.1 geohot 2025-02-05 09:08:28 +08:00
• a51c688f39 Cleanup llvm cleanup (and some clang things too) (#8871) uuuvnandGitHub 2025-02-05 01:49:05 +02:00
• bb5ded85cc Don't rewrite idiv to rshift when numerator is negative (#8885) eliotgoldingandGitHub 2025-02-04 23:47:33 +00:00
• 666b6149bc Use full soname for libgcc_s in CPUProgram (#8642) (#8896) 3 people 2025-02-05 00:45:48 +01:00
• 48349efdc1 copy is already contiguous (#8886) chenyuandGitHub 2025-02-04 17:53:33 -05:00
• 4c28235bd1 am: remove hardcodes (#8895) nimlgenandGitHub 2025-02-05 00:52:53 +03:00
• 057c70b05f add onnx_helpers to extra and add ort validate to benchmark_onnx (#8890) 3 people 2025-02-05 05:36:01 +08:00
• 89eebd4bfb pow cleanups (#8894) chenyuandGitHub 2025-02-04 15:52:57 -05:00
• 7a9e3247c2 simple start to the Kernel UOp [pr] (#8893) qazalandGitHub 2025-02-04 21:48:15 +01:00
• b4e8878e01 remove tensor_uops tracking from ScheduleContext [pr] (#8892) qazalandGitHub 2025-02-04 20:34:15 +01:00
• 6a0da51ed0 truncate process replay logs [pr] (#8891) qazalandGitHub 2025-02-04 20:26:48 +01:00
• c7c279a6bd unbind ShapeTrackers without maintaining a cache [pr] (#8889) qazalandGitHub 2025-02-04 19:43:41 +01:00
• 61de654efa minor shard cleanup [pr] (#8888) chenyuandGitHub 2025-02-04 13:22:31 -05:00
• 6ec7f1b00f replace UPat(name="x") with UPat.var("x") [pr] (#8887) qazalandGitHub 2025-02-04 19:12:40 +01:00
• c26b06eaeb delete fold_img_cast [pr] (#8875) qazalandGitHub 2025-02-04 18:43:45 +01:00
• acf0baefee process replay from tensor uops to kernel ast (#8883) qazalandGitHub 2025-02-04 18:09:20 +01:00
• dcf104ee68 ptx wmma render refactor (#8873) 3 people 2025-02-04 13:01:23 -03:00
• b92f36179d don't use set in schedule + add GroupOp.All [pr] (#8882) qazalandGitHub 2025-02-04 08:19:27 +01:00
• 56fa5c1191 dsp simulator (#8869) George HotzandGitHub 2025-02-04 09:45:04 +08:00
• 836cf42c2e fix rand_like for multi (#8880) chenyuandGitHub 2025-02-03 19:00:14 -05:00
• 746d899dbd move multi axis to property (#8879) chenyuandGitHub 2025-02-03 16:02:09 -05:00
• fa90079370 amd: reallocate scratch (#8872) nimlgenandGitHub 2025-02-03 23:21:37 +03:00
• ec447a31e7 factor out get_axis in multi [pr] (#8878) chenyuandGitHub 2025-02-03 14:39:08 -05:00
• cce26009f0 simplify pow to not call cos (#8877) chenyuandGitHub 2025-02-03 12:54:18 -05:00
• d1aa9f30bc copy onnx_ops into onnx (#8876) geohotstanandGitHub 2025-02-04 01:15:07 +08:00