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

• 91eb93f85a ops_ext to replace cpu import (#3406) George HotzandGitHub 2024-02-15 12:14:58 +01:00
• 49cb1fee54 run test_indexing on remu (#3404) qazalandGitHub 2024-02-15 12:52:40 +02:00
• 9d4d63fcfc dynamic tc function render (#3387) qazalandGitHub 2024-02-15 12:19:46 +02:00
• 3c4c8c9e16 bump db version to 11 (#3398) chenyuandGitHub 2024-02-14 10:13:18 -05:00
• 27f4de2ce4 delete half_prekernel (#3388) qazalandGitHub 2024-02-14 16:40:48 +02:00
• 078a2603d5 set metal fast math default to 0 (disabled) (#3370) chenyuandGitHub 2024-02-14 05:42:33 -05:00
• 668324d92b wmma: protect TC locals from modification and use only LOCAL (#3379) Francis LamandGitHub 2024-02-13 01:19:35 -08:00
• f1ad01fd91 test_linearizer_failures: add new linearizer compile failure on METAL (#3380) Francis LamandGitHub 2024-02-12 17:28:34 -08:00
• ce1f9f5556 hotfix: new linearizer docs geohot 2024-02-12 18:56:30 +01:00
• 2e60012bcf move create schedule and delete old API (#3377) George HotzandGitHub 2024-02-12 18:10:45 +01:00
• 41efaa848c move graph.py and jit.py into features (#3376) George HotzandGitHub 2024-02-12 17:34:34 +01:00
• 0f6cde243d import from wino_cleanup (#3374) George HotzandGitHub 2024-02-12 16:26:50 +01:00
• f47e297d4e refactor: END -> ENDLOOP geohot 2024-02-12 15:46:18 +01:00
• 29d68ae637 uops endif (#3372) George HotzandGitHub 2024-02-12 15:43:37 +01:00
• 1d45f3899d use is instead of == (#3371) George HotzandGitHub 2024-02-12 15:35:55 +01:00
• 323393b650 verbose apply_matrix (#3333) 3 people 2024-02-12 03:06:12 -08:00
• d55f99e881 patch merge_views (#3311) Jyotirmaya MahantaandGitHub 2024-02-12 02:53:55 -08:00
• b6a2600c86 fix merging condition in merge_dims (#3363) Jyotirmaya MahantaandGitHub 2024-02-12 02:50:26 -08:00
• c8fd66a131 Run RDNA3 tensor core tests in CI (#3367) qazalandGitHub 2024-02-12 02:54:06 +02:00
• f798b60338 add METAL_FAST_MATH env var to disable metal fast math (#3369) chenyuandGitHub 2024-02-11 04:26:09 -05:00
• 1156a27619 cleanup atol in test_ops (#3368) chenyuandGitHub 2024-02-10 19:44:44 -05:00
• 98c732cf9d fix metal compile error in extra/gemm (#3365) Yoshinori SanoandGitHub 2024-02-10 20:54:41 +09:00
• d1fb1e0ba4 full sync to fix HIP memory leak (#3364) George HotzandGitHub 2024-02-10 11:50:27 +01:00
• ddb22a60c8 linearizer: fix up edge case bugs in UNROLL opt (#3362) Francis LamandGitHub 2024-02-10 02:49:25 -08:00
• dc82ef6660 hotfix: swap HIP/CUDA bringup order to prevent delay on tinybox geohot 2024-02-09 18:41:25 +01:00
• 28ba1c5406 fix Tensor.randint ignoring kwargs (#3350) andresgitandGitHub 2024-02-09 18:12:16 +02:00
• ce21fdfb67 ops_python: add HIP tensor core mock and refactor METAL (#3354) 3 people 2024-02-09 03:46:06 -08:00
• b385234961 oops, change to 3.12 (#3357) George HotzandGitHub 2024-02-09 12:21:06 +01:00
• 7726eef464 ops_python: add image support (#3356) George HotzandGitHub 2024-02-09 12:02:06 +01:00
• 5f93061f67 ops_python: gated load support (#3355) George HotzandGitHub 2024-02-09 11:16:25 +01:00
• c151131d1b update onnx tests that no longer fail on CI (#3353) chenyuandGitHub 2024-02-08 21:19:00 -05:00
• 7c1c6efee5 exclude half with PYTHON in test_dtype.is_dtype_supported (#3351) chenyuandGitHub 2024-02-08 20:10:25 -05:00
• c32ea95d7d Python uop emulator (#3327) George HotzandGitHub 2024-02-08 19:24:55 +01:00
• 3ebf7a3e38 reflect changes to shapetracker in doc printouts (#3349) Mason MahaffeyandGitHub 2024-02-08 10:20:30 -05:00
• 2266152b28 linearizer: added FUZZ_BEAM to fuzz_linearizer and additional tests (#3340) Francis LamandGitHub 2024-02-08 07:12:58 -08:00
• b110c4a7b8 explicitly set input low and high in test_ops (#3347) chenyuandGitHub 2024-02-08 04:11:45 -05:00
• d8ad9e5660 verify eval acc for hlb_cifar training (#3344) chenyuandGitHub 2024-02-07 19:19:59 -05:00
• 0d2dacb549 test intermediate tensors created by function have same device as input (#3338) chenyuandGitHub 2024-02-07 09:24:36 -05:00
• 1732f1ba83 fix import and long lines in view (#3337) chenyuandGitHub 2024-02-07 06:50:21 -05:00
• 02636ff62d re-enable test_reduce_0d_default int test case in test_dtype (#3336) chenyuandGitHub 2024-02-07 05:30:14 -05:00
• ca66be6a70 add failed Tensor.pow test cases (#3334) chenyuandGitHub 2024-02-07 04:28:24 -05:00
• ea74856d99 remove some noqa: E501 in tensor (#3332) chenyuandGitHub 2024-02-07 00:03:05 -05:00
• 6478ee5c75 PoC UnaryOps before expand (#3319) David HouandGitHub 2024-02-06 19:05:13 -08:00
• d9ef8e25b3 fix Tensor.var with 0 in reduce dim. (#3324) chenyuandGitHub 2024-02-05 20:59:13 -05:00
• ee25f73283 Fix Tensor.mean to compute the mean correctly when 0-length axes are selected (#3318) Obada KhaliliandGitHub 2024-02-05 08:40:37 +02:00
• 3752e97c8f Fix: Always cast ONNX Slice op arguments into ints (#3317) 3 people 2024-02-05 01:40:48 +02:00
• 97275101e9 fix safetensor load uint32 and uint64 (#3315) chenyuandGitHub 2024-02-04 10:46:27 -05:00
• edb74897b2 support safe load bf16 (#3310) Yoshinori SanoandGitHub 2024-02-05 00:08:39 +09:00
• ca7973f61c clean up einsum_mulacc (#3312) chenyuandGitHub 2024-02-04 06:21:19 -05:00
• d459956966 move TestGetContraction to test_helpers (#3313) chenyuandGitHub 2024-02-04 06:05:01 -05:00
• b4ea0e18e3 Fix dot product on buffers with zero strides (#3303) Obada KhaliliandGitHub 2024-02-04 12:15:06 +02:00
• 30a3288c4a touchup canonicalize empty mask (#3308) chenyuandGitHub 2024-02-03 21:05:10 -05:00
• f5e0d9673c canonicalize empty masks (#3292) Jyotirmaya MahantaandGitHub 2024-02-04 06:57:57 +05:30
• 7e6f69e963 remove some contiguous and contiguous_backward from wino (#3306) chenyuandGitHub 2024-02-02 23:15:05 -05:00
• f8563a7e9f touchup apply_matrix (#3301) chenyuandGitHub 2024-02-02 05:13:37 -05:00
• 3a7c1eb383 add winograd hlb_cifar10 back to tinybox benchmark (#3300) chenyuandGitHub 2024-02-02 04:29:56 -05:00
• aebaab011f faster wino compile by catting consts across data expand dim (#3293) David HouandGitHub 2024-02-02 00:47:45 -08:00
• cf6f478901 limit group_for_reduce bufs to 32kb (#3299) David HouandGitHub 2024-02-02 00:13:12 -08:00
• b564660637 type annotation for Compiler.cachekey and minor cleanup (#3298) chenyuandGitHub 2024-02-01 21:31:21 -05:00
• 021eea3a52 fix UnboundLocalError when running Compiler with DISABLE_COMPILER_CACHE (#3296) Felix WuandGitHub 2024-02-02 03:12:33 +01:00
• a5bf4afc1a update ruff.toml for v0.2.0 (#3297) chenyuandGitHub 2024-02-01 20:50:20 -05:00
• 9196b11dfb test_ops sinh/cosh/asinh/acosh/atanh (#3294) chenyuandGitHub 2024-02-01 03:10:11 -05:00
• 927f2dd24d wmma: add HIP FP16 to FP16 tensor core (#3287) Francis LamandGitHub 2024-01-31 20:00:51 -08:00
• 18e854cdbf shrink MLB on sharded axis (#3255) chenyuandGitHub 2024-01-31 21:48:25 -05:00
• a3652e6ddc minor cleanups to test_ops (#3290) chenyuandGitHub 2024-01-31 19:01:25 -05:00
• 77251336d5 fix handcode_resnet50_opt.py (#3289) chenyuandGitHub 2024-01-31 19:01:08 -05:00
• 9b8c1a0408 Tensor.batchnorm works more than 2d and reuse in onnx (#3284) chenyuandGitHub 2024-01-30 19:02:45 -05:00
• 7816c3b692 onnx update for trilu and argmax (#3283) chenyuandGitHub 2024-01-30 18:39:16 -05:00
• 5b46b0ff3d Simple RDNA3 emulator (#2974) 3 people 2024-01-30 20:39:28 +02:00
• 247a8a2a6c add canonicalization to View.create (#3280) George HotzandGitHub 2024-01-30 10:26:48 -08:00
• d8f6280ffb hotfix: add CHECK_NEQ to fuzz_shapetracker_math geohot 2024-01-30 10:07:54 -08:00
• 09f2952dc3 reintroduce merge views in update benchmark (#3279) George HotzandGitHub 2024-01-30 09:47:20 -08:00
• d298916232 Revert "take merge views from corsix branch" (#3278) George HotzandGitHub 2024-01-30 09:34:28 -08:00
• b57a16aa89 take merge views from corsix branch (#3273) George HotzandGitHub 2024-01-30 09:25:16 -08:00
• 6a4a5dc79d fix pad 0 size (#3277) George HotzandGitHub 2024-01-30 08:58:10 -08:00
• b0a755288f cifar EVAL_BS set default value to BS (#3274) chenyuandGitHub 2024-01-29 17:37:12 -05:00
• 861d5ac224 wmma: fix the upcasts after WMMA to be hcopt ordering invariant (#3250) Francis LamandGitHub 2024-01-29 11:51:57 -08:00
• af4ca85594 MultiLazyBuffer.reshape new_axis without real_strides (#3272) chenyuandGitHub 2024-01-28 23:53:52 -05:00
• 34c7621556 HIP=1 NOCLANG=1 for tinybox external_model_benchmark (#3270) chenyuandGitHub 2024-01-28 22:05:26 -05:00
• 085dc87bed winograd should be 4 kernels (#3268) George HotzandGitHub 2024-01-28 09:21:26 -08:00
• f48b6aca77 long running beam pool (#3267) George HotzandGitHub 2024-01-28 08:06:03 -08:00
• 9e17378b60 Fix metal tests (#3266) George HotzandGitHub 2024-01-27 18:09:42 -08:00
• 86748f4a8c fix bbox format to be a list (#3265) Francis LataandGitHub 2024-01-27 20:54:19 -05:00
• 67a78615e5 uoptimizer (#3262) George HotzandGitHub 2024-01-27 10:26:04 -08:00
• 3ae811af21 tests for Tensor init data dtype and resulting dtype (#3247) 3 people 2024-01-27 10:13:42 +02:00
• 3c728d1082 compiler support (#3260) George HotzandGitHub 2024-01-26 23:36:40 -08:00
• 4273aabe31 extra/gemm: add a simple_conv.py along with correctness check (#3236) Francis LamandGitHub 2024-01-26 19:06:57 -08:00
• 0aad8d238b rebuild ocelot (#3259) George HotzandGitHub 2024-01-26 18:46:36 -08:00
• 473935125a use comgr to compile (#3248) George HotzandGitHub 2024-01-26 18:27:49 -08:00
• c4d870db0d fix jit realize issue (#3258) George HotzandGitHub 2024-01-26 18:27:35 -08:00
• 4197ef17c4 const cleanup with dtype.Scalar (#3257) chenyuandGitHub 2024-01-26 21:16:22 -05:00
• 03a6bc59c1 move autogen to runtime/autogen (#3254) George HotzandGitHub 2024-01-26 12:44:19 -08:00
• a3869ffd46 move gpuctypes in tree (#3253) George HotzandGitHub 2024-01-26 12:25:03 -08:00
• bc92c4cc32 onnx Einsum, CumSum, DepthToSpace, SpaceToDepth (#3252) chenyuandGitHub 2024-01-26 10:47:53 -05:00
• e45ffdb6cf cleanup onnx (#3249) chenyuandGitHub 2024-01-25 20:39:59 -05:00
• 168b1f879c Fix hip_matmul gemm in extra (#3241) Ahmed HarmoucheandGitHub 2024-01-26 01:03:04 +01:00
• 7feeb118e6 hip launch speed (#3246) George HotzandGitHub 2024-01-25 15:13:55 -08:00
• cb372b053f add device speed test (#3244) George HotzandGitHub 2024-01-25 12:01:22 -08:00
• d0e116c6d6 fix maximum/where Scalar casting (#3194) geohotstanandGitHub 2024-01-26 01:26:04 +08:00
• 3628bea910 fix: big round even rounder round (#3242) geohotstanandGitHub 2024-01-26 01:24:15 +08:00