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

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

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

delta_attention

delta_attention_scan

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

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

gl_dims

gldims

glm_flash

grad_outerworld

grok_x86_novec

group_div_rules

hcq2_vram_copies

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_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

no_weakint_variable

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

outer_range

outerworld_work

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

qwen_mergable

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_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

vf_ish

view_after_const

view_in_codegen

viz_cbroadcast

warp_fun

wgpu-f16-packed

working_postopt

x86_moves

x86_numel

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

• 026ba78526 Add commit hooks (#478) Jacky LeeandGitHub 2023-01-26 22:24:31 -08:00
• c07bc39941 fix mypy, plz add commit hooks George Hotz 2023-01-26 14:25:42 -08:00
• f08e740957 factor out hand coded opt Comma Device 2023-01-26 14:54:06 -06:00
• 5e8a36a18b real op kernel George Hotz 2023-01-26 09:51:32 -08:00
• e0600f537a op kernel in kernel search George Hotz 2023-01-26 09:47:01 -08:00
• 60acb2641f ugh, don't use os geohot 2023-01-25 19:41:21 -08:00
• b1dec64815 new types and fixup ShapeTracker type mismatches geohot 2023-01-25 19:39:36 -08:00
• 1b624a5051 DeviceBuffer has abstract methods geohot 2023-01-25 19:16:23 -08:00
• faab6461dd that lambda is required geohot 2023-01-25 18:46:56 -08:00
• 44e96c58b4 touch up pytorch speed tests geohot 2023-01-25 18:11:26 -08:00
• 8db345d846 functools.partialmethod -> lambda fixes Python 3.11 geohot 2023-01-25 18:08:08 -08:00
• a0af1045bf Some new tests (#440) calleditandGitHub 2023-01-26 00:40:19 +01:00
• aafc29484a cleanups geohot 2023-01-25 12:37:10 -08:00
• 919e943867 decent search geohot 2023-01-25 12:20:53 -08:00
• 7f3da91f8b kernel_search geohot 2023-01-25 12:05:09 -08:00
• e37424424f first little attempt at search geohot 2023-01-25 11:49:29 -08:00
• c15e9c3c7a comment where future perf should go geohot 2023-01-25 11:13:57 -08:00
• ee1f6ab3ca flip output shape extra dimension indexing for speed geohot 2023-01-25 11:00:37 -08:00
• 335a261a2e test for slow kernel geohot 2023-01-25 10:25:22 -08:00
• 0d594ccc51 mps option in torch (note: it's broken) geohot 2023-01-25 10:10:31 -08:00
• 66da3bc3c0 reset the benchmark timer geohot 2023-01-25 09:20:34 -08:00
• f5be4043ac fix OSX CL kernel timing geohot 2023-01-25 08:37:18 -08:00
• f6fc2a0d98 huh, this prevents an extra kernel geohot 2023-01-25 07:53:35 -08:00
• 487685919b Revert "Rename Normalize and move to nn (#415)" (#474) George HotzandGitHub 2023-01-25 07:50:04 -08:00
• d768acb6a9 Rename Normalize and move to nn (#415) Jacky LeeandGitHub 2023-01-25 07:47:59 -08:00
• baf64c14ac cleanups, simple padding in the processing op geohot 2023-01-25 07:37:52 -08:00
• 3acf62d489 cleanups for IMAGE=2 conv geohot 2023-01-25 07:18:34 -08:00
• 6d7658db12 delete opencl <celebration> George Hotz 2023-01-24 14:18:35 -08:00
• e313c8af20 update openpilot tests from OPENCL to GPU George Hotz 2023-01-24 14:05:20 -08:00
• 2e1d47b166 there's a bug in scc for empty string George Hotz 2023-01-24 12:06:06 -08:00
• e9c293361b fix typo George Hotz 2023-01-24 12:03:58 -08:00
• 9e2af0a972 too far with the OPTWG Comma Device 2023-01-24 13:14:59 -06:00
• 3590848b93 a little more local workgroup options Comma Device 2023-01-24 12:50:27 -06:00
• 4b74752c42 fix hotspots by improving the workgroup optimizer Comma Device 2023-01-24 12:46:28 -06:00
• fd760a390a fix incremental time George Hotz 2023-01-24 10:19:04 -08:00
• 7a369b856b nope, no default NATIVE_EXPLOG George Hotz 2023-01-24 10:01:52 -08:00
• 78fedc13d1 native_explog is default geohot 2023-01-24 08:09:43 -08:00
• 5d350d4883 the ast test is actually a test now geohot 2023-01-24 07:53:24 -08:00
• 7a159b9b04 tinygrad got big...make it tiny again geohot 2023-01-23 21:33:56 -08:00
• 6286ace4f1 does this work yet (#471) George HotzandGitHub 2023-01-23 20:36:17 -08:00
• c22554f44a floats for nvidia geohot 2023-01-23 16:36:10 -08:00
• 6fe9edf30f torch cuda is very fast geohot 2023-01-23 16:24:46 -08:00
• a949de873b reduce 2.0 (#469) George HotzandGitHub 2023-01-23 15:11:13 -08:00
• a6de94b444 test partial sum geohot 2023-01-22 21:28:40 -08:00
• f1196984e6 harmless to intertwine the math and the stores geohot 2023-01-21 09:31:56 -08:00
• 708215d06b Typing (#468) George HotzandGitHub 2023-01-21 09:09:22 -08:00
• b29614592a first conv/second conv George Hotz 2023-01-19 13:26:11 -08:00
• 3d697577b2 print_ast George Hotz 2023-01-19 13:22:03 -08:00
• 325a440cb5 pass in op_estimate in opencl George Hotz 2023-01-19 11:02:23 -08:00
• 844c645834 add flops for processing op George Hotz 2023-01-19 10:58:44 -08:00
• 0881d504c1 move shapetracker (#466) George HotzandGitHub 2023-01-19 09:56:31 -08:00
• 2b47ee401f Symbolic for indexes (#464) George HotzandGitHub 2023-01-19 07:21:30 -08:00
• 3a3400e3a2 more from indexer geohot 2023-01-18 18:11:51 -08:00
• 9245f4650a indexer changes for master geohot 2023-01-18 18:02:02 -08:00
• 15d04f13ce expr_idxs geohot 2023-01-15 10:41:59 -08:00
• 70b771a175 idx idy geohot 2023-01-15 09:39:22 -08:00
• 7ea89779fa add returns between views geohot 2023-01-15 08:58:10 -08:00
• 287699c32c simplify ones after axis splitting geohot 2023-01-14 10:51:43 -08:00
• 1b5def5b9d flip image x/y to match OPENCL geohot 2023-01-12 17:45:37 -08:00
• 49c6e6d472 Latest attempt to add image (#462) George HotzandGitHub 2023-01-12 17:36:30 -08:00
• 698143aebc shapetracker creates less views geohot 2023-01-12 14:04:04 -08:00
• 281b0db773 three from image geohot 2023-01-12 12:26:58 -08:00
• 7824ba256e no fake contiguous geohot 2023-01-11 21:27:53 -08:00
• 795285ce43 actually fix early alloc geohot 2023-01-11 21:12:50 -08:00
• 0724fd61f1 fix early alloc geohot 2023-01-11 21:07:27 -08:00
• f1378b3ea1 fix linter and force allocation on hostbuf geohot 2023-01-11 21:04:21 -08:00
• 3ea38cac72 IMAGE == 1, add reshape to the ast geohot 2023-01-11 20:56:03 -08:00
• 9ff6c532eb Prereqs for IMAGE=1 (#461) George HotzandGitHub 2023-01-11 20:18:42 -08:00
• 74fb772e2a rawcpu is junk, replace with llvm geohot 2023-01-10 19:18:45 -08:00
• fff1f046b0 Simple version of the new GPU backend (#458) George HotzandGitHub 2023-01-10 19:16:02 -08:00
• 66123c99b9 gpubuffer repr match llvmbuffer geohot 2023-01-09 20:02:22 -08:00
• 0a7d2b1a2e fix kernel_cnt type geohot 2023-01-09 19:34:57 -08:00
• 4356683081 gpu: rename kernels geohot 2023-01-09 19:32:22 -08:00
• 1e1abb450e fromcpu geohot 2023-01-09 19:18:57 -08:00
• 90121482fa oops, don't assign self geohot 2023-01-09 18:02:12 -08:00
• fad7cba590 move batchnorm to Tensor geohot 2023-01-09 18:00:16 -08:00
• 27211103ae docker: no -it geohot 2023-01-09 12:49:59 -08:00
• d6e86a29a8 docker: forgot to checkout code geohot 2023-01-09 12:48:03 -08:00
• 73ce9a771e that fix it geohot 2023-01-09 12:46:33 -08:00
• bfd4f4e35c testdocker geohot 2023-01-09 12:41:44 -08:00
• 4885fce56e shapetracker from newgpu (#456) George HotzandGitHub 2023-01-09 12:40:01 -08:00
• 538b1d7f5b Print out the tensor using numpy(). (#454) Faisal MemonandGitHub 2023-01-09 18:08:05 +00:00
• 2e744ef2f2 confirmed (#449) nogiraandGitHub 2023-01-08 03:41:06 +11:00
• 8dbf76268d Add step for setting up Stable Diffusion (#452) Nicolai StoianovandGitHub 2023-01-07 17:40:12 +01:00
• 4fb97b8de0 don't fail when termcolor is not installed (#436) cloud11665andGitHub 2022-11-15 01:45:06 +01:00
• 5e07d4669d the speedy chonker is going to replace the old chonker (#432) George HotzandGitHub 2022-11-11 18:34:24 -08:00
• d2273d2cc4 s/contiguous_op/contiguous geohot 2022-11-11 00:07:05 -08:00
• b8c94a67c9 Simple chonker (#431) 3 people 2022-11-10 23:17:09 -08:00
• bff47e9dc1 contiguous, and no strided for matmul geohot 2022-11-09 16:56:26 -08:00
• 1271f19a2b factorizing shapetracker from chonker geohot 2022-11-09 16:36:31 -08:00
• 64ff1ddc10 Reduce line count (#424) Daniel DavisandGitHub 2022-11-09 13:07:22 -05:00
• 0994705166 contrib more geohot 2022-11-08 19:14:37 -08:00
• c0bba9649a more that geohot 2022-11-08 19:13:11 -08:00
• 5143da6a9f contributing geohot 2022-11-08 19:12:05 -08:00
• 4998bf49b3 Basic editorconfig support (#422) Daniel DavisandGitHub 2022-11-08 13:34:25 -05:00
• c3d9c9b24c Fix issue where batch_invstd not being set (#421) marcojobandGitHub 2022-11-08 18:24:53 +01:00
• 8dc28dd733 Create python-publish.yml (#163) v0.4.0 LiamandGitHub 2022-11-08 17:45:01 +01:00
• 92ed87b0a5 bump version to 0.4.0 geohot 2022-11-08 08:44:42 -08:00
• 9781b4c3af rename test functions to helper_ geohot 2022-11-07 21:27:56 -08:00
• 9884be2ad5 ugh, that too geohot 2022-11-07 21:21:35 -08:00