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

• 7b7b9c1ae0 spec tests for movement ops before custom_kernel (#17299) qazalandGitHub 2026-07-30 18:20:24 +08:00
• aba5ba4b40 update rangeify comment for mop after AFTER (#17298) qazalandGitHub 2026-07-30 16:09:32 +08:00
• 3af1d62571 tests passing geohot 2026-07-30 03:24:48 +00:00
• d52ef3077c fix do_devectorize dtype (#17295) chenyuandGitHub 2026-07-29 23:04:25 -04:00
• bfc9fc6e0a nicer TinyJit decorator (#17293) sirhcmandGitHub 2026-07-29 21:48:00 -04:00
• f0295493a8 Merge remote-tracking branch 'origin/master' into llm_262k_minimal geohot 2026-07-30 00:51:47 +00:00
• 6cd7cc0888 unused geohot 2026-07-29 23:27:06 +00:00
• 060f447db6 qcom: match cl for SP_CS_INSTR_SIZE (#17289) sirhcmandGitHub 2026-07-29 18:47:02 -04:00
• fd912b348c generic c0*x<c1 [pr] (#17291) chenyuandGitHub 2026-07-29 18:45:23 -04:00
• 138676ab81 improve fragment example + index unshard (kimi) (#17288) George HotzandGitHub 2026-07-29 15:38:38 -07:00
• 027907a544 fix c0*x<c1 symbolic [pr] (#17290) chenyuandGitHub 2026-07-29 18:32:31 -04:00
• d79daa6acb all kernels in tinygrad uop language and fast geohot 2026-07-29 22:30:59 +00:00
• d4ba8b6e0f hcq2: use stack (#17286) nimlgenandGitHub 2026-07-29 22:09:34 +03:00
• b30c7e00d4 support 2d on UNSHARD (kimi) (#17285) George HotzandGitHub 2026-07-29 12:01:59 -07:00
• aab51fb7b6 64-bit UOp.variable support, try 2 (#17256) sirhcmandGitHub 2026-07-29 14:39:36 -04:00
• dd86a30798 test case for weakfloat cast to weakint INDEX (#17287) chenyuandGitHub 2026-07-29 14:15:52 -04:00
• 52c9e5a99e rename LOOP -> WEAK and STRONGLOOP -> LOOP (#17283) George HotzandGitHub 2026-07-29 10:38:36 -07:00
• 6c2b9fac08 weak const cleanups [PR] (#17282) chenyuandGitHub 2026-07-29 13:14:20 -04:00
• bd296a7359 enable alloc_fragment support with UNSHARD (kimi) (#17272) George HotzandGitHub 2026-07-29 09:46:45 -07:00
• 3df1b07c86 Patch half precision ops for older gpu archs (#17274) Noah SchiroandGitHub 2026-07-29 12:42:24 -04:00
• 3803f1583b fix add_raw_barrier [pr] (#17278) chenyuandGitHub 2026-07-29 12:30:00 -04:00
• 6ea7d366fa test permuted input in custom Ops.PROGRAM test (#17279) qazalandGitHub 2026-07-29 17:29:05 +08:00
• 527e57300c fix smu reset for kernel >= 7 (#17277) George HotzandGitHub 2026-07-28 23:11:21 -07:00
• 1e10302055 match kernel amd_smu_reset_patch geohot 2026-07-28 22:59:28 -07:00
• 599245fc64 fix smu reset for kernel >7 geohot 2026-07-28 22:56:14 -07:00
• e684fcc689 pm_reduce_collapse fix for re enabling stack for cat of same shape (codex) (#17251) qazalandGitHub 2026-07-29 13:47:28 +08:00
• 451120c6e1 make .barrier implicit (kimi) (#17275) George HotzandGitHub 2026-07-28 22:34:57 -07:00
• 2f8f2d2d37 remove no-op explicit dtype= or cast [PR] (#17276) chenyuandGitHub 2026-07-29 00:47:35 -04:00
• dd16d5aead apply shrink bugfix for 3.11 (#17271) George HotzandGitHub 2026-07-28 19:03:06 -07:00
• 291ee435bb qwen3.6 for 27b and 35b-a3b (#17268) chenyuandGitHub 2026-07-28 21:11:49 -04:00
• a17387d6a0 add UNSHARD to spec (#17269) George HotzandGitHub 2026-07-28 17:00:21 -07:00
• 57ae1bc7a7 rename MULTI to UNSHARD (#17267) George HotzandGitHub 2026-07-28 16:51:41 -07:00
• 1757067fa6 add device range as src[1] to multi (kimi) (#17264) George HotzandGitHub 2026-07-28 16:39:34 -07:00
• f11f884f49 update dtypes.md (#17266) chenyuandGitHub 2026-07-28 19:30:13 -04:00
• 23e9e76e8c DEFAULT_FLOAT/DEFAULT_INT ContextVar [pr] (#17265) chenyuandGitHub 2026-07-28 19:08:18 -04:00
• 755dfb243b rename CPU_COUNT to NUM_CPU_THREADS with cgroup awareness (#17263) 3 people 2026-07-28 15:33:57 -07:00
• 0cdddf3495 remove unneeded default args in renderers (#17262) George HotzandGitHub 2026-07-28 14:53:04 -07:00
• fde3a8faff CAPTURE_PROCESS_REPLAY=0 default chaging test [PR] (#17261) chenyuandGitHub 2026-07-28 17:32:35 -04:00
• 9ce65b71c0 delete dsp_pm_late (#17260) George HotzandGitHub 2026-07-28 14:05:41 -07:00
• 7d48926293 switch _device_num to AxisType.DEVICE range (kimi) (#17252) George HotzandGitHub 2026-07-28 13:52:47 -07:00
• d1c3ae0f91 preserve typed ranges through reshape (#17259) drkandGitHub 2026-07-28 13:51:06 -07:00
• 749e002286 remove some dtype= when construct UOp [pr] (#17257) chenyuandGitHub 2026-07-28 16:05:35 -04:00
• c9e11544df delete explicit casts [pr] (#17255) chenyuandGitHub 2026-07-28 13:57:13 -04:00
• a9ad08064e make the github actions runners generic for gitea (#17254) George HotzandGitHub 2026-07-28 10:48:49 -07:00
• 37cf15942b UOp.const(dtype=None) infers from from_py [pr] (#17253) chenyuandGitHub 2026-07-28 12:06:55 -04:00
• 0e1a2709f8 cpu: 134/36.6 geohot 2026-07-28 15:05:17 +00:00
• 05e4727feb arange stack regression test (#17250) qazalandGitHub 2026-07-28 19:00:22 +08:00
• e3b3eea1e2 test for extra copy in allreduce_cast (#16740) qazalandGitHub 2026-07-28 18:02:54 +08:00
• 1380d6cc5a pm_long_decomp doesn't depend on operand dtype [pr] (#17249) chenyuandGitHub 2026-07-28 01:15:04 -04:00
• 4b7022e8f4 Revert "64-bit UOp.variable support (#17246)" (#17248) chenyuandGitHub 2026-07-28 00:00:26 -04:00
• ab8fb191b2 64-bit UOp.variable support (#17246) sirhcmandGitHub 2026-07-27 23:56:09 -04:00
• f837ca3587 don't match const dtype in UPat [pr] (#17247) chenyuandGitHub 2026-07-27 22:50:08 -04:00
• 654d475d37 update pm_long_decomp [pr] (#17245) chenyuandGitHub 2026-07-27 22:28:50 -04:00
• 550225f603 clean up pm_lower_weak [pr] (#17243) chenyuandGitHub 2026-07-27 19:11:39 -04:00
• 0bb36c9989 make Tensor(None) weakfloat (#17241) chenyuandGitHub 2026-07-27 16:45:28 -04:00
• 896afad9bf don't match strong typed const in UPat [pr] (#17240) chenyuandGitHub 2026-07-27 16:19:01 -04:00
• 8b9ef157d1 run_linear in external_test_gpu_crash (#17239) nimlgenandGitHub 2026-07-27 22:00:11 +03:00
• bdbb1d702f fix shard axis through symbolic reshape (#17238) b1tgandGitHub 2026-07-27 23:32:15 +08:00
• 8eaeede96d hcq2: do not cache beam (#17234) nimlgenandGitHub 2026-07-27 17:16:24 +03:00
• 818a892ebc flip from_py to use weak dtypes [pr] (#17229) chenyuandGitHub 2026-07-27 10:07:00 -04:00
• a3ca1e55a5 polish the viz readme (#17236) qazalandGitHub 2026-07-27 21:03:13 +08:00
• 056974468e gptoss: split no-wd params (#17233) wozeparrotandGitHub 2026-07-27 02:58:43 -07:00
• 19c4d736f2 validate json output of viz.cli in CI (#17232) qazalandGitHub 2026-07-27 14:44:24 +08:00
• f45fc4c566 test updates from weak flip (#17231) chenyuandGitHub 2026-07-27 01:27:42 -04:00
• 95e3b0066f webgpu failing test for duplicate PARAM in CALL [pr] (#17230) qazalandGitHub 2026-07-27 13:06:06 +08:00
• 0f98212e80 skip test_float_to_fp8e4m3_extreme_values (#17228) chenyuandGitHub 2026-07-26 21:01:51 -04:00
• 456b5b5060 fix python_alu inf (#17227) chenyuandGitHub 2026-07-26 20:04:12 -04:00
• afb25a624c simpler minimum and copysign [PR] (#17226) chenyuandGitHub 2026-07-26 19:54:22 -04:00
• 165f0626f8 speedy hcq2 (#17225) nimlgenandGitHub 2026-07-27 01:31:56 +03:00
• 97a2265362 hcq2: amd indirect (#17220) nimlgenandGitHub 2026-07-26 21:36:01 +03:00
• 4b6760539b fix _prepare_jit_inputs for weak [pr] (#17222) chenyuandGitHub 2026-07-26 14:03:27 -04:00
• 94dad3d261 clean up mixin cos and exp [PR] (#17221) chenyuandGitHub 2026-07-26 13:13:05 -04:00
• a8d51097dc realize weak is no-op [pr] (#17219) chenyuandGitHub 2026-07-26 11:52:56 -04:00
• b434b17f90 cpu is 100/20 faster than llama.cpp geohot 2026-07-26 15:44:12 +00:00
• 79c07a334c fix ValueError in UOp.axis for shard reshape crossing boundary (#16547) 3 people 2026-07-26 21:08:10 +05:30
• d70a134845 fix nvrtc_check helper used for jitlink call (#16362) C TandGitHub 2026-07-26 18:20:39 +03:00
• 960430a5e5 Revert "nv: set lower interleave level to reduce GPU hogging (ai slop) (#15518)" (#17218) George HotzandGitHub 2026-07-26 08:15:16 -07:00
• ac12914506 nv: set lower interleave level to reduce GPU hogging (ai slop) (#15518) 4 people 2026-07-26 08:12:05 -07:00
• 5d1aa84901 SHR/SHL are Broadcastable [pr] (#17216) chenyuandGitHub 2026-07-26 02:11:03 -04:00
• acc2374b6f minor normalize cleanup [pr] (#17215) chenyuandGitHub 2026-07-26 00:39:53 -04:00
• a96974e70c slight weak behavior tweak and cleanups [pr] (#17214) chenyuandGitHub 2026-07-25 22:51:24 -04:00
• eb889053bf weak frompy prerequisite [PR] (#17213) chenyuandGitHub 2026-07-25 22:35:45 -04:00
• a60b5f77ac fix torch backend out= into a view (#17210) chenyuandGitHub 2026-07-25 21:04:42 -04:00
• 3946df787d hcq2, cpu is hcq2-ish (#17197) nimlgenandGitHub 2026-07-26 02:38:15 +03:00
• 076b37e1ae failing batch norm test (#17206) chenyuandGitHub 2026-07-25 18:46:49 -04:00
• 1a523193dd Merge branch 'master' into invalid_try_3 invalid_try_3 George HotzandGitHub 2026-07-25 14:07:16 -07:00
• e70a79465e invalid try 3 (kimi) geohot 2026-07-25 14:06:29 -07:00
• 492dc6d5fb delete _index_to_concrete_int [pr] (#17205) chenyuandGitHub 2026-07-25 16:05:08 -04:00
• 74c2121d99 promo (uint64, int) -> weakfloat like JAX [pr] (#17204) chenyuandGitHub 2026-07-25 15:20:42 -04:00
• f902513355 derive torch backend dispatch from the aten schema (#17203) chenyuandGitHub 2026-07-25 14:21:38 -04:00
• ee2ccb1f24 put weakint in dtypes.weaks [pr] (#17202) chenyuandGitHub 2026-07-25 14:00:34 -04:00
• f0117e98df refactor mlperf optim (#17200) wozeparrotandGitHub 2026-07-25 10:37:11 -07:00
• 983ad3bd95 fix torch backend batchnorm backward (#17201) chenyuandGitHub 2026-07-25 13:15:09 -04:00
• 9f78504304 checked cast in torch backend unwrap (#17199) chenyuandGitHub 2026-07-25 12:34:08 -04:00
• 732e6bd52c add one line viz mention (#17198) qazalandGitHub 2026-07-25 23:18:41 +08:00
• 8a10892f5a fix torch backend as_strided (#17195) chenyuandGitHub 2026-07-25 02:24:17 -04:00
• 4c58b260fb less wrong calculate_storage_offset (#17194) chenyuandGitHub 2026-07-25 01:53:54 -04:00
• d923263a5b simpler lower_weakint_node [PR] (#17193) chenyuandGitHub 2026-07-25 00:31:04 -04:00
• 9fdaa4bff2 standardize Program class (#17189) sirhcmandGitHub 2026-07-25 00:17:38 -04:00
• 0a6125e571 Invalid is bool, put weakint in promo lattice [pr] (#17188) chenyuandGitHub 2026-07-25 00:06:53 -04:00