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

• 4ee114b42a llama rope freqs in fp32 (#17698) qazalandGitHub 2026-08-24 11:49:55 +08:00
• 9d0cd0ebcb hotfix: switch benchmark to qwen3.8 geohot 2026-08-23 20:34:30 -07:00
• 11edcc144c disk cache: thread-local db conn (#17694) YassineYousfiandGitHub 2026-08-23 19:18:47 -07:00
• a31aca9e52 don't hardcode dtype.int const in decode_hevc_frame (#17696) chenyuandGitHub 2026-08-23 22:00:41 -04:00
• 8b164aefea test update from weak const (#17689) chenyuandGitHub 2026-08-23 17:30:16 -04:00
• 477b573807 update llm kv cache to be half (#17690) George HotzandGitHub 2026-08-23 14:02:54 -07:00
• bb0e99acbf hotfix: skip that nan test on mac geohot 2026-08-23 08:40:46 -07:00
• 93865e2c66 hcq2: sunday housekeeping (#17686) nimlgenandGitHub 2026-08-23 17:20:40 +03:00
• 5b60a09ab0 some fixes for the AMD emulator (#17684) George HotzandGitHub 2026-08-22 22:48:07 -07:00
• b0a1285330 lil decomp cleanup [PR] (#17682) chenyuandGitHub 2026-08-22 18:11:14 -04:00
• a2e64e16aa hcq2: early usb (#17683) nimlgenandGitHub 2026-08-23 00:22:39 +03:00
• a9069c177a make decomp pass SPEC=2 [PR] (#17681) chenyuandGitHub 2026-08-22 08:01:15 -04:00
• 8950942e75 remove explicit dtype for NOOP and decomp [PR] (#17678) chenyuandGitHub 2026-08-21 22:27:36 -04:00
• 356f665377 test update for weak const (#17675) chenyuandGitHub 2026-08-21 21:50:29 -04:00
• 7204d46786 delete dtype_from_uop INDEX exempt (#17674) chenyuandGitHub 2026-08-21 21:08:06 -04:00
• af242819d8 refactor the AMD emulator slop (kimi) (#17673) George HotzandGitHub 2026-08-21 18:00:58 -07:00
• 52596dbf38 gptoss: fused ce (#17672) wozeparrotandGitHub 2026-08-21 19:19:28 -04:00
• 07cce78cec compile3: log printed timings (#17671) sirhcmandGitHub 2026-08-21 16:18:32 -07:00
• f986829461 keep IndexingContext scoped in indexing (#17670) George HotzandGitHub 2026-08-21 14:12:39 -07:00
• 4fd4eafb23 nv: hevc (#17661) nimlgenandGitHub 2026-08-21 23:37:09 +03:00
• daa154aa22 FLOORDIV to SHR for powers of 2 [pr] (#17669) chenyuandGitHub 2026-08-21 16:28:04 -04:00
• 12f889aaad hcq2: parallel compile (#17667) nimlgenandGitHub 2026-08-21 23:27:21 +03:00
• 298748ebd3 ci: remove setup-python (#17665) sirhcmandGitHub 2026-08-21 13:19:47 -07:00
• 3082956a17 usb copyin: async arm and drain, 323 MB/s on comma (#17663) George HotzandGitHub 2026-08-21 13:12:03 -07:00
• 402bea7ddd l2i and sign_extend cleanups [pr] (#17668) chenyuandGitHub 2026-08-21 15:55:45 -04:00
• 8f9cbdf0cc few more self folding [pr] (#17657) chenyuandGitHub 2026-08-21 15:21:28 -04:00
• 8f59041ee5 more wgsl pack cleanups [PR] (#17664) chenyuandGitHub 2026-08-21 14:34:25 -04:00
• 1cf8a2c7fe hcq2: use shrink.bitcast (#17653) nimlgenandGitHub 2026-08-21 21:24:34 +03:00
• 3919ce8427 ceildiv in wgsl _packed_size [pr] (#17662) chenyuandGitHub 2026-08-21 14:02:20 -04:00
• 756e82e055 usb amd: pipelined copyin over the 0xF2 engine (2.6x faster) (#17628) 3 people 2026-08-21 10:38:50 -07:00
• cc32aa18db don't match casted const in const_folding_pat [pr] (#17658) chenyuandGitHub 2026-08-21 12:16:07 -04:00
• 77f698e55b cleanup the mxfp4 gemm (#17660) qazalandGitHub 2026-08-21 23:59:04 +08:00
• 554d078ac4 fix render_marg [pr] (#17656) chenyuandGitHub 2026-08-21 10:03:22 -04:00
• 176377ff6e weak 1 for FDIV in get_late_rewrite_patterns [PR] (#17655) chenyuandGitHub 2026-08-21 09:33:43 -04:00
• 1c3c9e96f6 remove setUp from test_sqtt_profiler (#17652) qazalandGitHub 2026-08-21 16:29:11 +08:00
• e8a8d99b99 profiler tracing using unique keys (#17651) qazalandGitHub 2026-08-21 16:06:01 +08:00
• dcc2d021e7 prereq viz cleanups for unique profile keys (#17649) qazalandGitHub 2026-08-21 15:18:25 +08:00
• 80bf60d782 profiler failing test for non unique kernel names (#17647) qazalandGitHub 2026-08-21 12:52:35 +08:00
• 1cb0600086 fix llm vars regression (kimi) (#17644) George HotzandGitHub 2026-08-20 21:31:43 -07:00
• 1bcb6bdc62 no-op weak.py reorder [PR] (#17646) chenyuandGitHub 2026-08-20 23:56:06 -04:00
• d716d0d927 hotfix: qkv bw kernel requires hipcc and cdna4 (#17645) qazalandGitHub 2026-08-21 11:38:45 +08:00
• 9216aa494c llm prefill failing test (#17630) b1tgandGitHub 2026-08-21 09:58:34 +08:00
• 9aa9e11301 compile kernels in parallel (#17629) George HotzandGitHub 2026-08-20 16:41:45 -07:00
• 3fdbb82bfe add ansipad and PARALLEL contextvar (#17642) George HotzandGitHub 2026-08-20 16:26:02 -07:00
• 0ccef542e0 add extract support to fetch (#17641) George HotzandGitHub 2026-08-20 16:20:16 -07:00
• c655aaf3a2 ci: venv in /opt/venv (#17632) sirhcmandGitHub 2026-08-20 15:54:54 -07:00
• 592e3f8363 update const selector where folding [pr] (#17640) chenyuandGitHub 2026-08-20 18:49:34 -04:00
• 3715006a21 fix float_to_bf16 on non-float32 inputs (#17638) sirhcmandGitHub 2026-08-20 15:14:30 -07:00
• d80254c1d9 fa: remove swa recompute (#17635) wozeparrotandGitHub 2026-08-20 17:26:44 -04:00
• c773891e3f skip INT_MIN % -1 on X86 and LLVM (#17637) sirhcmandGitHub 2026-08-20 14:26:26 -07:00
• 0e7ab863a0 x86 REX issue (#17580) chenyuandGitHub 2026-08-20 16:49:04 -04:00
• a57188ea6d hotfix: disable HCQ2 geohot 2026-08-20 13:39:48 -07:00
• 1707dca3b4 remove kernel_cnt, names are no longer unique (#17633) George HotzandGitHub 2026-08-20 13:28:25 -07:00
• eba5b7e750 benchmarks: cleanups (#17631) sirhcmandGitHub 2026-08-20 13:12:14 -07:00
• a8ecb73363 x64 imm uint64 (#17636) chenyuandGitHub 2026-08-20 15:30:17 -04:00
• f55c1a37d2 clean up some index(dtype=) [PR] (#17634) chenyuandGitHub 2026-08-20 15:18:15 -04:00
• 6732d05157 fix dtype_from_uop for invalid ALU [PR] (#17626) chenyuandGitHub 2026-08-20 10:47:23 -04:00
• e68aa16e3f hcq2: faster beam (#17624) nimlgenandGitHub 2026-08-20 15:21:53 +03:00
• 8c3cb00d36 hcq2: staging (#17622) nimlgenandGitHub 2026-08-20 13:42:48 +03:00
• c117da9850 safe changes for mi350p (#17620) George HotzandGitHub 2026-08-20 00:00:23 -07:00
• 57d1104a92 bump amd firmware (#17621) George HotzandGitHub 2026-08-19 23:51:15 -07:00
• a1263fadf3 fused_qkv_rope in UOp try 2 (#17619) qazalandGitHub 2026-08-20 11:28:34 +08:00
• e6324d1e1c test updates from weak const branch (#17618) chenyuandGitHub 2026-08-19 22:56:25 -04:00
• c63d94e059 benchmarks: split multigpu (#17615) sirhcmandGitHub 2026-08-19 18:36:07 -07:00
• c89ae6c083 gptoss: save more (#17613) wozeparrotandGitHub 2026-08-19 17:57:18 -04:00
• 2067133732 cpu: link with rt (#17608) sirhcmandGitHub 2026-08-19 14:55:08 -07:00
• ab68c58759 hcq2: speed (#17604) nimlgenandGitHub 2026-08-19 23:02:17 +03:00
• fc214da417 test updates for weak const change (#17606) chenyuandGitHub 2026-08-19 15:56:01 -04:00
• 0a0b6cb596 fix TestDevCopySpeeds command (#17607) chenyuandGitHub 2026-08-19 15:54:41 -04:00
• b8cc74ecf8 no float in tensor shape [pr] (#17605) chenyuandGitHub 2026-08-19 15:40:16 -04:00
• 7064e76bc8 fix roll on zero-sized tensors (#17603) 3 people 2026-08-19 15:31:21 -04:00
• 0c5307b4f3 realize: fast stat (#17600) nimlgenandGitHub 2026-08-19 22:04:06 +03:00
• a4fadcf606 fix TestDevCopySpeeds SIZE (#17602) chenyuandGitHub 2026-08-19 14:53:17 -04:00
• c218b4842d fold_bitcast should truncate its input [pr] (#17601) chenyuandGitHub 2026-08-19 14:39:36 -04:00
• bd6e70ac15 delete stale tests (#17596) chenyuandGitHub 2026-08-19 11:18:05 -04:00
• 9550378704 finish casted_consts migration [PR] (#17595) chenyuandGitHub 2026-08-19 10:43:55 -04:00
• b3e2f17b24 update NULL tests that depends on strong dtype CONST (#17594) chenyuandGitHub 2026-08-19 10:26:38 -04:00
• e8ba214b56 casted CONST migration for x86 [pr] (#17592) chenyuandGitHub 2026-08-19 09:38:09 -04:00
• 68b4407fe3 casted CONST migration for cstyle [pr] (#17587) chenyuandGitHub 2026-08-19 09:01:40 -04:00
• d539aaf752 Revert "fused_qkv_rope in UOp (#17591)" (#17593) qazalandGitHub 2026-08-19 20:42:55 +08:00
• 8c2bf02d17 fused_qkv_rope in UOp (#17591) qazalandGitHub 2026-08-19 17:07:30 +08:00
• ca86a42703 casted CONST migration for nir [pr] (#17588) chenyuandGitHub 2026-08-18 23:04:07 -04:00
• df3b114fbc ci: standardize all ubuntu runs-on to ubuntu-24.04 (#17586) sirhcmandGitHub 2026-08-18 18:49:07 -07:00
• e37b44d048 casted CONST migration for LLVM and PTX [pr] (#17585) chenyuandGitHub 2026-08-18 21:07:43 -04:00
• 2cfb421a81 ci: cleanup deps (#17583) sirhcmandGitHub 2026-08-18 16:51:03 -07:00
• c31038ff37 use KernelCountException when kernel count is being compared (#17584) George HotzandGitHub 2026-08-18 16:06:03 -07:00
• c0d9c79668 add ninja to setup-tinygrad gitea-test-ci ci-deps sirhcm 2026-08-18 15:53:24 -07:00
• 19dc868750 ci: cleanup deps sirhcm 2026-08-18 15:48:48 -07:00
• 49778d9a48 start renderer casted const migration [pr] (#17582) chenyuandGitHub 2026-08-18 17:43:27 -04:00
• 72280bb218 gptoss: zero-2 optim (#17581) wozeparrotandGitHub 2026-08-19 05:28:57 +08:00
• af2a43c850 hcq2: 64bit addresses (#17576) nimlgenandGitHub 2026-08-18 16:52:18 +03:00
• a1366e2f6c alu(long, weakint) can do math in int too [pr] (#17579) chenyuandGitHub 2026-08-18 09:08:35 -04:00
• 0b757bb9bc Revert "disk: neable polling (#17538)" (#17578) nimlgenandGitHub 2026-08-18 15:31:25 +03:00
• 7cbe8e0d15 viz: expanding srcs should not override history (#17577) qazalandGitHub 2026-08-18 16:45:43 +08:00
• a746861ac0 compile server for cuda on mac (#17574) sirhcmandGitHub 2026-08-17 19:55:59 -07:00
• 8d2cc64b69 llm: refactor delta attention (#17564) George HotzandGitHub 2026-08-17 19:24:03 -07:00
• cb892e1b92 base_rewrite reorder [PR] (#17575) chenyuandGitHub 2026-08-17 20:55:44 -04:00
• d4a1f39038 clean up cstyle render inf and nan [PR] (#17573) chenyuandGitHub 2026-08-17 18:01:11 -04:00
• 34c9b9d434 add back cast where rule [pr] (#17572) chenyuandGitHub 2026-08-17 15:27:20 -04:00
• 901d257a26 some more torch backend cleanups (#17571) chenyuandGitHub 2026-08-17 15:19:57 -04:00