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

• b79454ba57 Deployed 0640cfc with MkDocs version: 1.6.1 gh-pages github-actions[bot] 2026-08-28 22:37:57 +00:00
• d9a2cad39c ci: set DEV explicitly explicit-dev sirhcm 2026-08-28 15:32:13 -07:00
• 0640cfce35 remove spec dtype check that's trivially true [PR] (#17823) master chenyuandGitHub 2026-08-28 17:00:24 -04:00
• 086f45cb81 dtype is not a UOp field anymore (#17821) chenyuandGitHub 2026-08-28 16:17:29 -04:00
• cb0db2d042 dtype_from_uop is never None (#17819) chenyuandGitHub 2026-08-28 15:28:09 -04:00
• 793c1664f3 don't set UOp dtype in x86 (#17818) chenyuandGitHub 2026-08-28 15:18:41 -04:00
• bd2711b7c1 put dtype to Ops.INS arg (#17817) chenyuandGitHub 2026-08-28 15:05:17 -04:00
• b3dad4fa0c update X86 cast related stuff (#17816) chenyuandGitHub 2026-08-28 14:30:34 -04:00
• 42bca39af9 test LP_NUM_THREADS=4 test-lp sirhcm 2026-08-28 11:16:43 -07:00
• 49627aecb5 ci: use fetch to download libraries (#17814) sirhcmandGitHub 2026-08-28 10:53:19 -07:00
• 959958135d am: vf (#17802) nimlgenandGitHub 2026-08-28 20:48:59 +03:00
• f196af2327 fix onnx MeanVarianceNormalization arg (#17812) chenyuandGitHub 2026-08-28 13:47:47 -04:00
• 112e50ec3a fix onnx.py type annotations (#17811) chenyuandGitHub 2026-08-28 13:34:43 -04:00
• ae63f23557 fix onnx MaxUnpool (#17809) chenyuandGitHub 2026-08-28 13:13:47 -04:00
• 287679a88a delete X86 not used codes [PR] (#17807) chenyuandGitHub 2026-08-28 12:50:50 -04:00
• 2eacd4fa68 fix Conv2d same padding with per axis dilation (#17805) Teddy TennantandGitHub 2026-08-28 12:44:43 -04:00
• 7fdc58b1cc add support for UD quants (#17806) Joel LisenbyandGitHub 2026-08-28 09:38:34 -07:00
• cb857a5a40 remove X86 ymm thingy (#17803) chenyuandGitHub 2026-08-28 10:18:10 -04:00
• 6083de5cff fix nan compares (#17801) chenyuandGitHub 2026-08-28 09:17:18 -04:00
• 55c8e18930 hcq2 spec2 (#17798) nimlgenandGitHub 2026-08-28 12:51:55 +03:00
• 76dce1eb8d tiny hcq2 changes (#17797) nimlgenandGitHub 2026-08-28 11:29:00 +03:00
• 6242b09066 cleaner mxfp4 gemm prelude (#17796) qazalandGitHub 2026-08-28 14:59:35 +08:00
• 92d92c70a9 gptoss: faster grad handling (#17795) wozeparrotandGitHub 2026-08-28 01:06:33 -04:00
• 26c4b6319e fix interpolate with align_corners and an output size of 1 (#17790) Teddy TennantandGitHub 2026-08-27 22:17:05 -04:00
• 53cad325c7 copy movement op tests construct copies (#17792) qazalandGitHub 2026-08-28 10:09:11 +08:00
• 4456e62343 ci: skip apt steps if packages are already installed (#17794) sirhcmandGitHub 2026-08-27 19:06:08 -07:00
• 39d7fd0981 axis arg is dead (#17793) George HotzandGitHub 2026-08-27 18:41:09 -07:00
• 4e6bdac412 hotfix: disable HCQ2 geohot 2026-08-27 18:22:56 -07:00
• bc485d271f fix CMPLT between float16 and const 0 (#17791) chenyuandGitHub 2026-08-27 21:14:34 -04:00
• 6b361e4de7 move shape into arg for param/buffer (#17783) George HotzandGitHub 2026-08-27 18:13:24 -07:00
• 38e0fe103b fix where with float64 branches (#17789) chenyuandGitHub 2026-08-27 21:03:43 -04:00
• e6ad2907a8 onnxruntime respects NUM_CPU_THREADS (#17788) sirhcmandGitHub 2026-08-27 17:11:27 -07:00
• fede358811 fix fancy indexing with uint8 index (#17787) chenyuandGitHub 2026-08-27 19:59:27 -04:00
• 56189c1ec4 onnxruntime respects NUM_CPU_THREADS onnx-cpu-count sirhcm 2026-08-27 16:41:01 -07:00
• f06832bf6f test llm with --no_chat_template (#17785) sirhcmandGitHub 2026-08-27 16:01:55 -07:00
• 1babc5e7a9 test llm with --no_chat_template no-chat-template sirhcm 2026-08-27 15:04:49 -07:00
• ee3161e924 hotfix: decrease dims in test_attention to get below the 90s limit geohot 2026-08-27 14:44:06 -07:00
• 38fa0643cf hotfix: update max test time from 300 -> 90 geohot 2026-08-27 14:27:22 -07:00
• f267a49639 UOp.param accepts single sint as shape (int,) (#17784) George HotzandGitHub 2026-08-27 14:07:33 -07:00
• 14cbecef8f autogen: use deb for comgr 2 (#17782) sirhcmandGitHub 2026-08-27 13:32:16 -07:00
• 793f3cf5a1 autogen: use deb for comgr 2 autogen-comgr-2 sirhcm 2026-08-27 12:02:13 -07:00
• 54d69f439f fix test_assign for no subbuffer backends (#17757) chenyuandGitHub 2026-08-27 10:47:21 -04:00
• 980c42d611 hcq2 tiny speedups (#17773) nimlgenandGitHub 2026-08-27 17:42:37 +03:00
• 7eb763a3c2 bnxt to extra (#17772) nimlgenandGitHub 2026-08-27 17:29:17 +03:00
• 5c3d044465 fix asinh gradient at zero (#17758) Teddy TennantandGitHub 2026-08-27 09:51:17 -04:00
• f19d89e29a hcq2: a bit cleaner deps (#17771) nimlgenandGitHub 2026-08-27 14:24:59 +03:00
• 2d4e7afcbf mockgpu/amd: use generated KFD ioctl metadata (#17761) tonyfloatersuandGitHub 2026-08-27 05:24:37 -04:00
• 0d66815849 set BEAM_PADTO=0 in llama (#17769) qazalandGitHub 2026-08-27 15:45:33 +08:00
• d16a2510ad remove some dead scheduler tests (#17768) qazalandGitHub 2026-08-27 15:01:28 +08:00
• 118e812b31 simplify more same device copies (#17766) qazalandGitHub 2026-08-27 11:54:12 +08:00
• 0820c7a424 benchmarks: openpilot matrix (#17714) sirhcmandGitHub 2026-08-26 15:53:55 -07:00
• 22ffff3258 gptoss: single jit (#17525) wozeparrotandGitHub 2026-08-26 18:32:21 -04:00
• a0a901c8e4 faster qwen 3.8 (#17720) George HotzandGitHub 2026-08-26 15:23:25 -07:00
• c015351ac5 fix _get_cpu_count for docker --cpus=N in python 3.13+ (#17760) sirhcmandGitHub 2026-08-26 13:53:30 -07:00
• 6074c002e1 hcq2: fix jit (#17747) nimlgenandGitHub 2026-08-26 22:36:36 +03:00
• 6042b87272 delete PCONTIG [PR] (#17756) George HotzandGitHub 2026-08-26 12:07:28 -07:00
• cc72b9f7be cleanup BENCHMARK_LOG (#17754) sirhcmandGitHub 2026-08-26 11:55:54 -07:00
• 6a3b297548 fix PTX NIR SPEC=2 for bool [pr] (#17753) chenyuandGitHub 2026-08-26 14:07:29 -04:00
• ea6c82f3be small changes from new rangeify (#17752) George HotzandGitHub 2026-08-26 10:33:31 -07:00
• 0abcf09b74 never bufferize_to_store weak input [PR] (#17751) chenyuandGitHub 2026-08-26 13:25:52 -04:00
• 4bdc865131 delete unused rewrite rules [PR] (#17748) chenyuandGitHub 2026-08-26 10:50:42 -04:00
• 4c20f1d357 fix asinh precision loss on negative inputs (#17749) Teddy TennantandGitHub 2026-08-26 10:08:04 -04:00
• ecf79e260d better all2all schedule test (#17746) qazalandGitHub 2026-08-26 14:46:10 +08:00
• 9860e5d285 llm tokenizer: fix tekken, add gpt4o (#17733) b1tgandGitHub 2026-08-26 14:30:35 +08:00
• 625c05df1e fix am_smi to respect dev (#17742) George HotzandGitHub 2026-08-25 19:21:55 -07:00
• b49c03fb1c benchmarks: don't use sudo on mac (#17740) sirhcmandGitHub 2026-08-25 18:06:04 -07:00
• dc04c7820e lil fixes from new_rangeify (#17741) George HotzandGitHub 2026-08-25 17:15:28 -07:00
• 6ece327cf3 CUSTOM arg is (str, dtype) [PR] (#17737) chenyuandGitHub 2026-08-25 20:15:06 -04:00
• 07268b724f fix: external_test_gpu_crash on python 3.14 (#17739) wozeparrotandGitHub 2026-08-25 18:43:30 -04:00
• 55032514ce pin onnxruntime==1.24.1 (#17738) sirhcmandGitHub 2026-08-25 14:53:15 -07:00
• 2824504f90 usb amd: yield between signal polls (#17712) YassineYousfiandGitHub 2026-08-25 14:11:53 -07:00
• 7fef98c86e benchmarks: remove usage of sudo from linux runners (#17735) sirhcmandGitHub 2026-08-25 13:21:08 -07:00
• b831ca62d9 Reapply "disk cache: thread-local db conn (#17694)" geohot 2026-08-25 13:20:15 -07:00
• d851aca9ae hcq2: fix copy/call in usb (#17732) nimlgenandGitHub 2026-08-25 19:03:19 +03:00
• 7dc8b666e7 wgsl cast before load for packed [pr] (#17731) chenyuandGitHub 2026-08-25 11:46:31 -04:00
• 023bfdb380 fix viz for hcq2 (#17729) nimlgenandGitHub 2026-08-25 18:23:22 +03:00
• 9f01775cf4 dtype_from_uop for THREEFRY and FDIV (#17727) chenyuandGitHub 2026-08-25 08:43:59 -04:00
• 9607787ce1 delete more explicitly set dtype to UOp [PR] (#17726) chenyuandGitHub 2026-08-25 08:32:45 -04:00
• ab79879613 bring back sqtt examples tests (#17725) qazalandGitHub 2026-08-25 16:00:18 +08:00
• a5678317c2 split rangeify to prepare.py (#17722) George HotzandGitHub 2026-08-24 19:12:13 -07:00
• 1d694dd700 remove more explicitly set dtype [PR] (#17721) chenyuandGitHub 2026-08-24 22:03:22 -04:00
• 65ca68567e const cleanups [PR] (#17719) chenyuandGitHub 2026-08-24 21:37:14 -04:00
• a7df1a1ace const are weak 3 [pr] (#17695) chenyuandGitHub 2026-08-24 20:32:57 -04:00
• d9004cff22 use native shifts in Payne Hayek instead of pow2 mul/divs (#17717) RaineandGitHub 2026-08-24 20:11:46 -04:00
• 6f87158d77 feat: bump version to 0.14.0 (#17716) wozeparrotandGitHub 2026-08-24 18:40:15 -04:00
• 021c015eb4 benchmarks: more aggressive timeouts (#17715) sirhcmandGitHub 2026-08-24 15:32:36 -07:00
• 2e7790f16c benchmarks: cleanup dsp benchmark (#17706) sirhcmandGitHub 2026-08-24 15:09:01 -07:00
• 8df3dac0ec remove redundant UOp dtype [PR] (#17713) chenyuandGitHub 2026-08-24 18:00:02 -04:00
• ed110993d3 mi350p: aqua vanjaram bringup for the raw PCI driver (#17693) George HotzandGitHub 2026-08-24 14:32:16 -07:00
• f4cc28824c delete GETTUPLE on UNSHARD rule [PR] (#17711) chenyuandGitHub 2026-08-24 17:25:28 -04:00
• 76bf6b7eec llm: add DEBUG=2 for weights loading (#17710) George HotzandGitHub 2026-08-24 14:09:34 -07:00
• 08fbd25f1a AGENTS: no amend commits geohot 2026-08-24 13:34:42 -07:00
• 5aabbb1991 fix Muon weight decay being a no-op (#17709) George HotzandGitHub 2026-08-24 13:33:10 -07:00
• 7a887e84b2 am: use gc regs for xgmi (#17707) nimlgenandGitHub 2026-08-24 22:42:35 +03:00
• 7b21ffac00 PYLITERAL has dtypes.void (#17705) chenyuandGitHub 2026-08-24 14:37:15 -04:00
• 9b481a4893 dsp: cleanup temp linker script (#17704) sirhcmandGitHub 2026-08-24 11:05:00 -07:00
• 9b0c65688d llama script changes from the speedups branch (#17701) qazalandGitHub 2026-08-24 20:10:02 +08:00
• 7e561fcb97 mergable fast RDNA3 Qwen 3.8 (#17512) George HotzandGitHub 2026-08-23 22:46:35 -07:00
• a5ea95d8b8 usb: don't fail async transfers on signal interruption (#17692) YassineYousfiandGitHub 2026-08-23 21:30:00 -07:00
• df528499ce Revert "disk cache: thread-local db conn (#17694)" geohot 2026-08-23 21:13:29 -07:00