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

• 74e6b6c9fc types (#2346) chenyuandGitHub 2023-11-17 18:49:24 -05:00
• d2c0035c73 add back as_strided, move rebuilt mops to extra (#2344) chenyuandGitHub 2023-11-17 14:34:30 -05:00
• 064034c42c hip free event + a bit faster cpu time (#2342) nimlgenandGitHub 2023-11-17 20:50:49 +03:00
• ad3d7428fa good line shaves in st and faster (#2343) chenyuandGitHub 2023-11-17 11:00:26 -05:00
• 652d2de256 wow how did i think that was okay (#2339) George HotzandGitHub 2023-11-16 21:21:11 -08:00
• 8e22c0d95c everything can jit now (#2338) chenyuandGitHub 2023-11-16 23:54:57 -05:00
• a8875bd770 add types to lazy (#2327) 3 people 2023-11-17 05:48:41 +01:00
• 1d5501594e force rebuild of ocelot (#2334) George HotzandGitHub 2023-11-16 20:44:14 -08:00
• 0d0c74bac9 Assert for memory allocation failures (#2337) imaoloandGitHub 2023-11-16 23:14:16 -05:00
• aa01a63b3f cleanup of lines / unused / types (#2336) chenyuandGitHub 2023-11-16 21:15:32 -05:00
• 3971259832 fix test_real_world llama (#2335) chenyuandGitHub 2023-11-16 19:50:08 -05:00
• 3b9dd3330c add device to beam search cache key (#2333) chenyuandGitHub 2023-11-16 18:35:08 -05:00
• 75676ab8e1 Profiling-helper (#2321) 3 people 2023-11-16 23:15:56 +01:00
• 8235da11dd whisper: support batch inference, add librispeech WER test (#2074) 3 people 2023-11-16 13:50:08 -08:00
• 3baaf298d6 two stage cumsum in tensor.py (#2331) George HotzandGitHub 2023-11-16 12:09:53 -08:00
• 163b2bc26a wgpu.utils._device -> wgpu.utils.device (#2330) chenyuandGitHub 2023-11-16 12:52:13 -05:00
• 27f4c26312 fix getitem slice when end < start (#2329) chenyuandGitHub 2023-11-16 11:20:27 -05:00
• 822d6e6f18 Simpler mops verify (#2325) chenyuandGitHub 2023-11-15 21:47:18 -05:00
• ef67d7ff5d shapetracker whitespace geohot 2023-11-15 15:24:09 -08:00
• a98511561c fuzz_linearizer same api for interpreted and compiled (#2320) chenyuandGitHub 2023-11-15 17:40:22 -05:00
• 294e71de15 remove lines (unused code) (#2319) George HotzandGitHub 2023-11-15 14:36:11 -08:00
• 628365eab6 JIT cleanups (#2317) George HotzandGitHub 2023-11-15 13:34:52 -08:00
• b64738e1d6 Remove AS_STRIDED from shapetracker (#2216) forcefieldsovereignandGitHub 2023-11-15 12:50:17 -08:00
• b8d460d203 Add Tensor.multinomial (#2295) Marcello FuschiandGitHub 2023-11-15 20:38:39 +01:00
• cb6cfcc8f8 add icb support check for metal device (#2313) taherandGitHub 2023-11-15 11:37:28 -08:00
• 70a65c201e JIT support in Interpreted (#2314) George HotzandGitHub 2023-11-15 11:13:38 -08:00
• 9a20bc08d6 Tensor(None) is Tensor([]) (#2316) chenyuandGitHub 2023-11-15 13:49:18 -05:00
• f1f863c953 allow 0-dim array to broadcast into zero shape tensor (#2315) chenyuandGitHub 2023-11-15 13:12:21 -05:00
• 4da2ddea6e Interpreted cleanups (#2312) George HotzandGitHub 2023-11-15 09:02:23 -08:00
• 123a0b86b2 support zero in shape (#2303) chenyuandGitHub 2023-11-15 11:57:48 -05:00
• f113a0b83b dtype promotion priorities (#2311) qazalandGitHub 2023-11-15 10:19:52 -05:00
• 3c5a51fb3a aaaaaaa finally (#2310) geohotstanandGitHub 2023-11-15 23:12:38 +08:00
• cff8375aa2 make self referential AST fast too (#2278) kormannandGitHub 2023-11-15 16:12:07 +01:00
• 4f7b1ac0d2 cleanups before interpreted jit (#2306) George HotzandGitHub 2023-11-14 21:44:25 -08:00
• 91546225f4 Add cache step for model weights in CI, re-enable whisper test (#2307) mmmkkaaayyandGitHub 2023-11-14 21:16:04 -08:00
• 175cdbe815 fix pad None will value (#2308) chenyuandGitHub 2023-11-14 23:57:05 -05:00
• 01f8781c26 fix CI (#2300) George HotzandGitHub 2023-11-14 11:02:59 -08:00
• 4e0d47533e beam works with var vals (#2296) nimlgenandGitHub 2023-11-14 21:03:19 +03:00
• fac8633ba8 explicit opts for test_linearizer_failures (#2299) chenyuandGitHub 2023-11-14 11:52:38 -05:00
• 8916028ddd move BatchExecutor (#2297) George HotzandGitHub 2023-11-14 08:08:51 -08:00
• 0cbf6c1811 move things, clean up extra (#2292) George HotzandGitHub 2023-11-13 20:18:40 -08:00
• b1f7f29525 metal indirect command buffers (#2285) George HotzandGitHub 2023-11-13 17:58:26 -08:00
• d86ea188dd support symbolic shape in Interpreted (#2289) chenyuandGitHub 2023-11-13 20:13:18 -05:00
• 6960bcded0 back to 6.54GB for stable diffusion (#2288) George HotzandGitHub 2023-11-13 16:50:04 -08:00
• 960535dfb8 get_linearizer_actions does not return illegal actions (#2287) nimlgenandGitHub 2023-11-13 19:48:54 +03:00
• 53c5baa8b6 add dilation to avg_pool2d (#2270) 5 people 2023-11-13 16:47:56 +00:00
• a72b370066 llama take int and convert to Variable internally (#2284) chenyuandGitHub 2023-11-12 17:11:37 -05:00
• 123ea051e6 refactor/ci: delete many # type: ignore (#2281) valarandGitHub 2023-11-12 11:04:20 -08:00
• 2e2154ae4f bad hotfix for optimize_local_size, try again geohot 2023-11-12 10:41:11 -08:00
• 270f747065 hotfix optimize_local_size (TODO: add regression test) geohot 2023-11-12 10:29:00 -08:00
• f5a62a1b42 fix some tests related to JitItem (#2279) chenyuandGitHub 2023-11-11 23:00:35 -05:00
• 5ef8d682e3 clean up attentions in stable diffusion (#2275) chenyuandGitHub 2023-11-11 14:25:36 -05:00
• 453f48ce02 pad None means (0,0) (#2273) chenyuandGitHub 2023-11-11 12:50:26 -05:00
• c5d70c1871 typo (#2271) jxdvandGitHub 2023-11-11 15:18:04 +00:00
• 880e693207 fix llama n_kv_heads in kvcache (#2267) chenyuandGitHub 2023-11-10 21:44:39 -05:00
• 78623ba204 two simple tests geohot 2023-11-10 16:16:06 -08:00
• 70fb8a259d hotfix mypy geohot 2023-11-10 15:43:30 -08:00
• 6ceea02e65 hotfix of onnx geohot 2023-11-10 15:40:30 -08:00
• b853e9bb8c Onnx 1.15.0 gogogo (#2217) geohotstanandGitHub 2023-11-11 07:36:48 +08:00
• 85d26ddc36 uops loop removal (#2262) George HotzandGitHub 2023-11-10 15:24:47 -08:00
• a753c8e071 examples of new GPT2 and JIT change (#2261) 3 people 2023-11-10 15:07:02 -05:00
• b6aaf12df7 Internal cast 2 with more tests (#2257) qazalandGitHub 2023-11-10 13:42:39 -05:00
• c0f447d6f7 Inline barrier (#2255) George HotzandGitHub 2023-11-10 08:17:10 -08:00
• 75f6e9ab54 one more fuzz linearizer failed example (#2260) chenyuandGitHub 2023-11-10 09:17:37 -05:00
• 330484c072 Revert "Internal casting support (#2046)" (#2256) George HotzandGitHub 2023-11-09 21:27:13 -08:00
• 7e1d08b2ae Internal casting support (#2046) 3 people 2023-11-10 00:02:32 -05:00
• 6051f0ce82 For cuda get current free space from device, and retry alloc failures (#2197) vish-prandGitHub 2023-11-10 05:23:50 +05:30
• 2465d5d267 fix ops tests in test_dtype (#2237) qazalandGitHub 2023-11-09 18:17:43 -05:00
• 80bf0b8586 proper wmma (#2245) 3 people 2023-11-09 15:15:18 -08:00
• b7a31fb708 remove tokei badge from readme (#2251) wozeparrotandGitHub 2023-11-09 13:53:31 -05:00
• 50bf0703aa fix sqlite cache path on Windows (#2250) 2-5andGitHub 2023-11-09 18:32:34 +00:00
• 10d642e174 fuzz linearizer transformation (#2188) chenyuandGitHub 2023-11-09 11:03:27 -05:00
• 794122781d Merge pull request #2242 from chenyuxyz/mypy-casts chenyuandGitHub 2023-11-08 20:04:46 -05:00
• 38b7f5a7fd less phi, proper phi (#2241) George HotzandGitHub 2023-11-08 16:13:43 -08:00
• b9fe133af8 mypy check warn_redundant_casts chenyu 2023-11-08 15:06:55 -08:00
• 4c44d1344b feat: remove cache_id (#2236) wozeparrotandGitHub 2023-11-08 11:09:21 -05:00
• 553688f12a update metal matmul and matvec for compile api (#2238) Rory ClearandGitHub 2023-11-08 16:08:35 +00:00
• 3042450b4d diskcache touchups (#2235) George HotzandGitHub 2023-11-07 18:00:04 -08:00
• 09bdd55acc update debug prints geohot 2023-11-07 17:47:25 -08:00
• c0a033f01d remove real_offset (#2234) George HotzandGitHub 2023-11-07 17:30:53 -08:00
• 4d95e6d070 move cache out of tmp (#2232) George HotzandGitHub 2023-11-07 11:41:00 -08:00
• a48ccdb359 cleanup deps, no pyyaml, pillow to testing (#2231) George HotzandGitHub 2023-11-07 10:32:23 -08:00
• ae5d1407ee Fix mmaped in jit (#2225) nimlgenandGitHub 2023-11-07 01:54:21 +03:00
• 0c9b4ab885 no to_underlying (#2222) George HotzandGitHub 2023-11-05 21:34:20 -08:00
• fbe7f0c62b metal: unwrap lib write geohot 2023-11-05 21:02:31 -08:00
• 2f7aab3d13 move optimize_local_size (#2221) George HotzandGitHub 2023-11-05 21:00:52 -08:00
• c60c3b467a clean up symlinking in benchmark (#2219) George HotzandGitHub 2023-11-05 16:46:05 -08:00
• baeb77a403 Make the JIT simple (no batch exec, no cache collector) (#2215) George HotzandGitHub 2023-11-05 16:23:43 -08:00
• 719a97b337 fix IMAGE=2 failed with NOOPT=1 (#2209) chenyuandGitHub 2023-11-05 16:16:37 -05:00
• 680cbfdba4 less broken limit_dims_to_max (#2214) chenyuandGitHub 2023-11-04 11:38:06 -04:00
• 265304e7fd Stable diffusion WebGPU port (#1370) 3 people 2023-11-04 02:29:16 +01:00
• f582ec56d5 Replace (getenv("CI", "") != "") with helpers.CI (#2213) chenyuandGitHub 2023-11-03 18:20:44 -04:00
• f17bc16f46 simple runtime args (#2211) George HotzandGitHub 2023-11-03 12:31:29 -07:00
• 9ea0448103 compile interpreted to python code (#2208) George HotzandGitHub 2023-11-03 09:16:12 -07:00
• ddbc6eecaf some refactors in the realization (#2206) George HotzandGitHub 2023-11-02 19:51:28 -07:00
• 51fd993f1f pin onnx to 1.14.1 geohot 2023-11-02 18:03:21 -07:00
• 6621d2eb98 Revert "Modernize setup.py (#2187)" George Hotz 2023-11-03 01:01:15 +00:00
• 6e06adcb95 fix hip segfault (#2204) nimlgenandGitHub 2023-11-02 18:40:56 +03:00
• 03cf0afa4f move all to compile api (#2203) George HotzandGitHub 2023-11-01 23:01:32 -07:00
• 8932816816 remove arm64, caching for cuda (#2201) George HotzandGitHub 2023-11-01 18:44:00 -07:00