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

• aa7e16744e allow sinking childless consts and fold them [pr] (#7941) qazalandGitHub 2024-11-28 07:23:37 -05:00
• 3ab67d45b2 init changes from the global buffer branch [pr] (#7939) qazalandGitHub 2024-11-28 06:38:58 -05:00
• 81d415be03 amd pkt3 refactor (#7923) nimlgenandGitHub 2024-11-28 11:06:37 +03:00
• e3fe7023b0 move all VIEW -> LOAD rules to big graph rewrite [pr] (#7936) qazalandGitHub 2024-11-28 01:02:29 -05:00
• e2eccdab43 swizzle upat consistency + assert it's base [pr] (#7935) qazalandGitHub 2024-11-28 00:35:55 -05:00
• c5c3b05b5a block lin: only the test changes (#7933) George HotzandGitHub 2024-11-28 13:19:00 +08:00
• 32dbab945c Revert "add block uops and modify tests (#7931)" (#7932) George HotzandGitHub 2024-11-28 13:15:41 +08:00
• 6f4519ff45 add block uops and modify tests (#7931) George HotzandGitHub 2024-11-28 13:11:18 +08:00
• 336a9b6bf3 remove dtype from llama precompute_freqs_cis (#7930) chenyuandGitHub 2024-11-27 22:28:40 -05:00
• 3e2430f822 use tqdm tqdm in mlperf training (#7929) chenyuandGitHub 2024-11-27 21:57:05 -05:00
• 864758423e Don't take const in gcd and change the "nothing_changed" condition (#7926) Sieds LyklesandGitHub 2024-11-28 00:07:36 +01:00
• 988d64900b add TODO case to test_mod_congruence (#7925) chenyuandGitHub 2024-11-27 15:23:21 -05:00
• 57262c8e34 update Tensor.scatter doc examples (#7924) chenyuandGitHub 2024-11-27 11:42:36 -05:00
• cea5853cfa add Tensor.scatter (#7737) 3 people 2024-11-27 23:52:04 +08:00
• 38f34ca0cb prepare mypy==1.13.0: legacy cast (#7866) JaSpa99andGitHub 2024-11-27 16:33:35 +01:00
• 753f07e193 add circular pad mode to Tensor.pad (#7918) 3 people 2024-11-27 23:30:51 +08:00
• a58e289d77 Revert "prereqs for new block lin so PR works (#7919)" (#7921) chenyuandGitHub 2024-11-27 08:41:09 -05:00
• c53261b541 prereqs for new block lin so PR works (#7919) George HotzandGitHub 2024-11-27 15:07:54 +08:00
• a6171cbe71 add stable diffusion v2 to mac benchmark (#7917) chenyuandGitHub 2024-11-26 22:09:43 -05:00
• d318867776 Factoring gcd out of mod (#7916) Sieds LyklesandGitHub 2024-11-27 03:17:22 +01:00
• 84f96e48a1 hcq signal tiny refactor (#7913) nimlgenandGitHub 2024-11-26 21:48:38 +03:00
• 345457f518 webgpu cache packages (#7911) qazalandGitHub 2024-11-26 11:17:36 -05:00
• 6102e3159c webgpu -n=auto (#7910) qazalandGitHub 2024-11-26 08:13:12 -05:00
• cab461c2b5 match lazy view in uop try 2 (#7905) qazalandGitHub 2024-11-26 07:31:50 -05:00
• ea57c52b99 base uop is always contiguous (#7907) qazalandGitHub 2024-11-26 07:13:29 -05:00
• ceda43ce75 always swizzle load st in wmma [pr] (#7908) qazalandGitHub 2024-11-26 07:00:58 -05:00
• 4e5bf9dc7a test assignment in jit (#7906) George HotzandGitHub 2024-11-26 17:37:00 +08:00
• 0cd1cc29dc PTX simplify: use a dict matcher for prefix [pr] (#7890) 3 people 2024-11-26 17:32:48 +08:00
• 10618aba98 Bring back WebGPU (#7063) 3 people 2024-11-26 05:26:40 +01:00
• ff3f2a9c1a Revert "move attention upcast (#7830)" (#7903) chenyuandGitHub 2024-11-25 18:59:51 -05:00
• 04bee97d2a hotfix ctypes.c_ulong(size) for metal _alloc (#7902) chenyuandGitHub 2024-11-25 18:25:33 -05:00
• 631dc98b52 validate llama quantize output (#7901) chenyuandGitHub 2024-11-25 16:46:23 -05:00
• e8777cb8db assert view on uops without shape [pr] (#7898) qazalandGitHub 2024-11-25 07:43:50 -05:00
• a49ca0c2ff clean up fully_flatten [pr] (#7885) 3 people 2024-11-25 06:53:18 -05:00
• e823de3828 viz with bottom_up=True (#7894) qazalandGitHub 2024-11-25 04:56:48 -05:00
• 2ca41d6a44 ops metadata map try 2, early fuse [pr] (#7893) qazalandGitHub 2024-11-25 04:08:38 -05:00
• 9295c86ddc delete base op cast [pr] (#7891) qazalandGitHub 2024-11-25 03:38:32 -05:00
• 26784c45c6 delete cast arg 2 [pr] (#7881) qazalandGitHub 2024-11-25 03:15:57 -05:00
• 9d0038bccb small changes from block linearizer [pr] (#7888) George HotzandGitHub 2024-11-25 15:27:04 +08:00
• 9e958f2b10 Ptx simplify [pr] (#7877) 3 people 2024-11-25 15:01:47 +08:00
• e9c681c839 fix missing final rewrite in viz (#7883) nib9888andGitHub 2024-11-25 06:13:33 +00:00
• a49a7c4784 Improved mod folding (#7887) Sieds LyklesandGitHub 2024-11-25 04:21:34 +01:00
• 5d92efb121 [BUGFIX] Tensor([]).data() (#7884) leopfandGitHub 2024-11-24 22:42:57 +01:00
• ac57d82a13 test_tiny on real NV/CUDA/AMD/HIP (#7886) chenyuandGitHub 2024-11-24 16:34:54 -05:00
• 06a28d83f5 delete extra dtype check in uop const [pr] (#7880) qazalandGitHub 2024-11-24 11:06:52 -05:00
• 31337b49e3 cleanup Embedding call [pr] (#7869) chenyuandGitHub 2024-11-24 07:32:26 -05:00
• ad9df26fba add test for inconsistent behavior in float to int casting (#7870) geohotstanandGitHub 2024-11-24 20:31:34 +08:00
• 6b8a657085 cleanup group_realizes [pr] (#7878) qazalandGitHub 2024-11-24 05:16:46 -05:00
• 5aee78a0a6 fix uop swizzle on BUFFER, new tests (#7875) qazalandGitHub 2024-11-24 04:11:09 -05:00
• 5d28a202b5 make tinychat local (#7871) George HotzandGitHub 2024-11-24 14:45:48 +08:00
• 22d5def113 download llama3 70B (#7868) chenyuandGitHub 2024-11-23 12:18:31 -05:00
• 6a8be3ca1e don't change lazy state in schedule [pr] (#7867) qazalandGitHub 2024-11-23 11:18:50 -05:00
• 28e83e662e least controversial (#7863) JaSpa99andGitHub 2024-11-23 14:23:30 +01:00
• 8c3d3181dd bottom up rewrite fixes substitute [pr] (#7862) George HotzandGitHub 2024-11-23 20:53:37 +08:00
• 54d8f75d0c vectorized define_acc does not seem to get used (#7858) 3 people 2024-11-23 19:46:34 +08:00
• 40be9177ba move swizzle upats to ops, prereq for swizzle tc [pr] (#7861) qazalandGitHub 2024-11-23 05:34:45 -05:00
• 27a6cd7822 cleanup swizzle upats [pr] (#7860) qazalandGitHub 2024-11-23 02:19:06 -05:00
• 5b2c03e865 defer realize folding to kernel splitting [pr] (#7849) qazalandGitHub 2024-11-23 01:29:14 -05:00
• 144e9f00df viz is local, new test, and new quantize [pr] (#7859) George HotzandGitHub 2024-11-23 14:27:10 +08:00
• d43613e113 refactor image cast folding [pr] (#7852) qazalandGitHub 2024-11-23 00:59:21 -05:00
• c07daf40e7 move attention upcast (#7830) chenyuandGitHub 2024-11-22 17:10:51 -05:00
• 5c5b1b994c less flaky benchmarks (#7855) chenyuandGitHub 2024-11-22 16:39:39 -05:00
• 3b26e51fce Tensor.cummax (#7854) chenyuandGitHub 2024-11-22 15:55:02 -05:00
• fb10ea563e typedef bf16 amd (#7850) ignaciosicaandGitHub 2024-11-22 16:29:01 -03:00
• a352a6938f simplify group_for_reduces in get_index [pr] (#7851) chenyuandGitHub 2024-11-22 11:53:21 -05:00
• af5d77f684 move sint_to_uop from view.py to ops.py [pr] (#7848) chenyuandGitHub 2024-11-22 11:15:02 -05:00
• f6d1201c48 variable_to_uop -> sint_to_uop [pr] (#7847) chenyuandGitHub 2024-11-22 10:54:59 -05:00
• 40d7535eeb clean up DTYPES_DICT [pr] (#7845) chenyuandGitHub 2024-11-22 10:01:34 -05:00
• 4453ab51e1 use ceildiv in View.stride [pr] (#7844) chenyuandGitHub 2024-11-22 08:38:05 -05:00
• 9828277c03 view doesn't have buffer, fix the tests [pr] (#7841) qazalandGitHub 2024-11-22 07:41:55 -05:00
• 7e8777eee9 faster assign scheduling [pr] (#7839) qazalandGitHub 2024-11-22 06:23:59 -05:00
• 6229d87f45 simpler reshape symbolic shape check [pr] (#7837) chenyuandGitHub 2024-11-21 22:53:57 -05:00
• 1d6d842887 move DSP to extra (room for webgpu) [pr] (#7836) George HotzandGitHub 2024-11-22 11:32:57 +08:00
• 8ff6cba9f0 simpler swizzle_r new_axis [pr] (#7835) chenyuandGitHub 2024-11-21 22:26:41 -05:00
• 6fc7013463 put all DSP in dsp file [pr] (#7833) George HotzandGitHub 2024-11-22 11:22:59 +08:00
• e39af63156 no loop assert in ops_python [pr] (#7834) George HotzandGitHub 2024-11-22 11:17:36 +08:00
• d18b948f48 ptxcompiler isn't a cudacompiler [pr] (#7832) George HotzandGitHub 2024-11-22 10:57:22 +08:00
• 855f9a767a add restype for msg method for type annotation and consistency (#7828) 3 people 2024-11-22 09:17:58 +08:00
• d5c9fafff5 default run stable diffusion benchmark with fp16 (#7831) chenyuandGitHub 2024-11-21 15:58:17 -05:00
• 69e382216d fix wino conv output dtype for half inputs (#7829) chenyuandGitHub 2024-11-21 12:13:54 -05:00
• cf1ec90ad4 add inverse trig functions to Tensor (#7805) 3 people 2024-11-21 22:13:36 +08:00
• 5399ff6d06 add UOp.const_with_shape [pr] (#7825) qazalandGitHub 2024-11-21 08:13:23 -05:00
• 2f884b2384 good suggestions from mypy lineprecision-report for schedule.py [pr] (#7823) qazalandGitHub 2024-11-21 06:59:51 -05:00
• e378aeb94e assert view degrade to const tests post scheduler graph_rewrite [pr] (#7822) qazalandGitHub 2024-11-21 06:00:41 -05:00
• cdc431803f early mark uops as realized [pr] (#7821) qazalandGitHub 2024-11-21 05:02:59 -05:00
• 4542c0f000 get buffer size just from Ops.BUFFER [pr] (#7820) qazalandGitHub 2024-11-21 04:00:18 -05:00
• 877b440fde derive device (dname) from UOp [pr] (#7819) qazalandGitHub 2024-11-21 03:38:22 -05:00
• 75c082b883 move CONST/BIND -> VALID to matchers (#7818) qazalandGitHub 2024-11-21 03:07:01 -05:00
• df6f1815ad remove jit_cache from self in GraphRunner [pr] (#7817) George HotzandGitHub 2024-11-21 13:26:37 +08:00
• e9ae2ccd09 _prg to match _buf [pr] (#7816) George HotzandGitHub 2024-11-21 12:44:48 +08:00
• 439911b2e6 disable disable_abstract_method [pr] (#7815) George HotzandGitHub 2024-11-21 12:28:57 +08:00
• c5d458ce02 BufferSpec and ProgramSpec [pr] (#7814) George HotzandGitHub 2024-11-21 12:18:05 +08:00
• 490a6130af more hcq typing [pr] (#7813) George HotzandGitHub 2024-11-21 11:23:07 +08:00
• 9df5a62c5e unify to HWQueue [pr] (#7812) George HotzandGitHub 2024-11-21 10:33:08 +08:00
• 11cea00090 lower vs_theoretical conv tflops threshold for nv (#7811) chenyuandGitHub 2024-11-20 20:03:49 -05:00
• 46aa23539f generate and print mypy lineprecision report (#7809) chenyuandGitHub 2024-11-20 16:53:17 -05:00
• c815d7b56e run bfloat16 tensor core in metal benchmark (#7808) chenyuandGitHub 2024-11-20 15:34:07 -05:00
• 33a496279b load_state_dict check v.shape instead of v.lazydata.shape (#7807) chenyuandGitHub 2024-11-20 14:39:30 -05:00
• fc3154a7b3 metal bf16 tc support [pr] (#7408) ignaciosicaandGitHub 2024-11-20 16:39:08 -03:00
• 66a069ee25 add replicate mode to Tensor.pad (#7802) geohotstanandGitHub 2024-11-20 21:39:58 +08:00