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

• 7ff7aacdb4 LazyOp out of Linearizer (#1908) George HotzandGitHub 2023-09-24 14:30:53 +08:00
• 2201b46bce Refactor Conv2d/ConvTranspose2d into a single parent class (#1906) qazalandGitHub 2023-09-24 09:23:41 +03:00
• 97dc813329 Revert "All LazyOps in the Linearizer (#1905)" (#1907) George HotzandGitHub 2023-09-24 11:51:22 +08:00
• a5820390db All LazyOps in the Linearizer (#1905) George HotzandGitHub 2023-09-24 11:50:00 +08:00
• 0f373b8b47 cache more uops (#1904) George HotzandGitHub 2023-09-23 16:50:13 +08:00
• 1e15fdaee7 disable flaky triton test geohot 2023-09-23 14:59:36 +08:00
• 0571dd7627 move all int (#1903) George HotzandGitHub 2023-09-23 14:43:45 +08:00
• 41aea3ad36 require C-contiguous array for hip._copyin (#1902) nimlgenandGitHub 2023-09-23 09:36:59 +03:00
• 58296c079d Make Triton work again (#1547) 4 people 2023-09-23 08:17:12 +02:00
• 6fb8b3bb60 move symbolic functions to shapetracker (#1901) George HotzandGitHub 2023-09-23 11:45:08 +08:00
• 9cf13bd055 rename reduce_op (#1900) George HotzandGitHub 2023-09-23 11:27:36 +08:00
• 73a6ed7862 Apply ShapeTracker in interpreted backends (#1846) George HotzandGitHub 2023-09-23 10:05:13 +08:00
• 3987280daf Fix VALIDHACKS for Images and make it default (#1832) Umut ZenginandGitHub 2023-09-23 02:34:43 +03:00
• 767bb35903 Enable symbolic ops tests for LLVM (#1898) Gijs KoningandGitHub 2023-09-23 01:30:26 +02:00
• b8ff20ffe4 Gpt2 (#1896) Gijs KoningandGitHub 2023-09-22 14:14:47 +02:00
• b89ee1ac83 lazy type annotation and cleanups (#1897) chenyuandGitHub 2023-09-22 02:20:23 -04:00
• 78576915de Add needed contiguous to DiskBuffer. SHM support on OSX (#1891) George HotzandGitHub 2023-09-22 09:16:42 +08:00
• d0e752003d fixes (#1893) qazalandGitHub 2023-09-22 02:20:27 +03:00
• 009a99a0b1 feat: way cleaner hip wrapper (#1895) wozeparrotandGitHub 2023-09-21 19:20:03 -04:00
• cb5d6576cb cifar step time 65ms while stay above 94% (#1888) Yixiang GaoandGitHub 2023-09-20 22:19:32 -05:00
• 864746d6aa polish print_tree (#1868) kormannandGitHub 2023-09-21 05:13:10 +02:00
• a5090f0ee9 remove NumNode.int() (#1876) chenyuandGitHub 2023-09-20 22:29:16 -04:00
• 9eb6310686 Fix gpt optimization (#1885) Gijs KoningandGitHub 2023-09-21 04:28:18 +02:00
• bd3444797b make ssa assign r[u] (#1887) Szymon OżógandGitHub 2023-09-21 04:20:20 +02:00
• 9450e41f70 no import when Python is shutting down (#1875) nimlgenandGitHub 2023-09-20 19:47:02 +03:00
• 84ab47a90a add branch up-to-date check (#1879) Yixiang GaoandGitHub 2023-09-20 11:41:51 -05:00
• 504bb6d0ea support symbolic jit in HIP (#1877) nimlgenandGitHub 2023-09-20 08:44:26 +03:00
• cd66c9e249 no numnode in shape (#1871) chenyuandGitHub 2023-09-16 16:49:45 -07:00
• 18ec5a9e09 add comment bot to CI (#1873) Yixiang GaoandGitHub 2023-09-16 11:22:06 -05:00
• a27f6c7d62 add diff mode to sz.py (#1872) Yixiang GaoandGitHub 2023-09-15 23:43:47 -05:00
• 4c31dfafb3 add seed to gpt-2 (#1869) nimlgenandGitHub 2023-09-16 00:34:14 +03:00
• c870764940 Revert "add line changes diff bot to CI (#1863)" (#1870) wozeparrotandGitHub 2023-09-15 16:56:42 -04:00
• 789c84a7a3 add line changes diff bot to CI (#1863) Yixiang GaoandGitHub 2023-09-15 15:29:58 -05:00
• 29ac8293d7 run gpt2 in CI (#1866) chenyuandGitHub 2023-09-14 13:37:02 -07:00
• 1b46de1a3e fix type of helpers.prod, add test cases (#1859) chenyuandGitHub 2023-09-13 14:16:55 -07:00
• e67306ba04 symbolic shape type with TypeGuard (#1852) chenyuandGitHub 2023-09-12 14:27:22 -07:00
• c91b44f7bf refactor: move size to view (#1848) 3 people 2023-09-11 16:16:04 +02:00
• 9e9ea20784 Fix view, CI cpu test with python 3.8 (#1845) chenyuandGitHub 2023-09-10 19:37:58 -07:00
• 3ec301c2d7 apply view.py patch (#1844) chenyuandGitHub 2023-09-10 17:32:15 -07:00
• a32951a001 add test_tensor_copy (#1840) Yixiang GaoandGitHub 2023-09-10 18:01:58 -05:00
• 1bc52c60df fix: minor tweaks to view (#1842) 3 people 2023-09-11 00:55:57 +02:00
• 47e602f717 view: do not trade complexity for speed (#1839) George HotzandGitHub 2023-09-10 11:29:53 -07:00
• c0bc4cfbaf DivNode.b is int (#1833) chenyuandGitHub 2023-09-10 09:04:29 -07:00
• 13790b1e20 cast types in render_load (#1837) nimlgenandGitHub 2023-09-10 17:58:13 +03:00
• e74a6ca7e4 expand in terms of substitute (#1827) David HouandGitHub 2023-09-09 14:43:00 -07:00
• 0e3e2bac13 amd wino: upload results geohot 2023-09-09 13:57:14 -07:00
• 6f95c5f284 winograd speed test for AMD (#1826) George HotzandGitHub 2023-09-09 13:56:33 -07:00
• 0f2bd10d00 add winograd CIFAR to mac tests (#1825) George HotzandGitHub 2023-09-09 13:45:24 -07:00
• 31fca43706 kopt works with local+grouped reduce and tests (#1824) nimlgenandGitHub 2023-09-09 23:22:09 +03:00
• 9da40c8448 move Node.__lt__ SumNode special case to SumNode (#1823) chenyuandGitHub 2023-09-09 13:20:38 -07:00
• 651205fa5c linearizer: support local and group_for_reduce dimensions together (#1821) Francis LamandGitHub 2023-09-08 12:39:27 -07:00
• 9e8c1dbf34 patch to remove hack from stable_diffusion.py (#1814) segf00ltandGitHub 2023-09-08 13:26:50 -03:00
• ebcda8a714 Move var_vals from ShapeTracker to LazyBuffer (#1819) chenyuandGitHub 2023-09-08 09:25:10 -07:00
• 7ac65a93b4 utils.printtree (#1816) kormannandGitHub 2023-09-08 08:08:57 +02:00
• 4613c9e77c add tvm example, formatting (#1813) George HotzandGitHub 2023-09-07 11:50:41 -07:00
• 5b15a972b5 no functions with same names in test/ (#1811) nimlgenandGitHub 2023-09-07 21:27:31 +03:00
• 722823dee1 stable diffusion: force fp16 free geohot 2023-09-06 15:11:05 -07:00
• 928cb1a64a AndNode.substitute short circuit (#1800) chenyuandGitHub 2023-09-06 14:58:49 -07:00
• a78a1fa499 fix jit buffer reuse when freed (#1802) nimlgenandGitHub 2023-09-07 00:41:57 +03:00
• 22cf15e9d0 convert function into tinygrad (#1803) Yixiang GaoandGitHub 2023-09-06 16:41:26 -05:00
• 52a92bf95d use class Foo: instead of class Foo(): (#1797) Pavol RusnakandGitHub 2023-09-06 21:20:25 +02:00
• fd25792c8b Ensure freqs as type float32 in freqs_cis (#1798) badccandGitHub 2023-09-06 10:24:15 -07:00
• 35072877ef sym_infer is noop for int input (#1795) chenyuandGitHub 2023-09-06 09:17:20 -07:00
• f67638b27a delete broken DDPG example geohot 2023-09-06 08:00:47 -07:00
• 78a43ad2c7 add uop fixup (#1793) George HotzandGitHub 2023-09-06 07:55:22 -07:00
• 1bbf26d7fd fix try except not catching fxn() in benchmark (#1783) geohotstanandGitHub 2023-09-06 22:36:43 +08:00
• 09e78a9d07 Node does not need to subclass ABC (#1792) chenyuandGitHub 2023-09-06 07:35:45 -07:00
• ee9ac20752 Use correct dtype in Tensor when data is an ndarray (#1785) badccandGitHub 2023-09-06 07:35:32 -07:00
• 130cd55942 fix gpu compilation of const GEP (#1788) nimlgenandGitHub 2023-09-06 17:34:46 +03:00
• e10a9692ec Revert "fix attn_mask None issue" (#1787) George HotzandGitHub 2023-09-05 21:18:55 -07:00
• 343b256deb PoC fast winograd compile (#1771) David HouandGitHub 2023-09-05 21:14:40 -07:00
• a50a7ef6f2 revert typo in external_multi_gpu.py (#1777) Pavol RusnakandGitHub 2023-09-06 05:46:28 +02:00
• bd06d88c73 fix attn_mask None issue (#1786) George HotzandGitHub 2023-09-05 20:45:54 -07:00
• 0379b64ac4 add seed option to stable_diffusion (#1784) Francis LamandGitHub 2023-09-05 19:45:15 -07:00
• 6100d7425f add 2 to locals, uops debug 5 (#1782) George HotzandGitHub 2023-09-05 19:44:43 -07:00
• 2a11669e1d perf: faster and more readable merge_dicts (#1775) 3 people 2023-09-05 23:42:19 +02:00
• 89a8a02697 disable openpilot model in model benchmark George Hotz 2023-09-05 13:32:30 -07:00
• 9af5645ba3 onnx full passing (#1076) 3 people 2023-09-06 04:23:32 +08:00
• fb1cc6bf4b llama jit is default, print tok/sec (#1774) George HotzandGitHub 2023-09-05 10:12:16 -07:00
• f6e6a1a4d7 perf: avoid cast, restore isinstance (#1772) 3 people 2023-09-05 15:07:04 +02:00
• 671101e6b8 Metal stuff pip install on default when on Darwin (#1770) geohotstanandGitHub 2023-09-05 12:59:54 +08:00
• 10305bfc0a tuples only (#1769) George HotzandGitHub 2023-09-04 16:35:11 -07:00
• 63c46e0287 Parens and gls (#1768) George HotzandGitHub 2023-09-04 16:09:01 -07:00
• 3473c9e88d Metal conv tensor cores (#1696) Adrian KretzandGitHub 2023-09-05 00:14:46 +02:00
• b32ed8e6e9 removing loop (#1764) George HotzandGitHub 2023-09-04 14:47:46 -07:00
• 7344f7c2d1 KeyError fixed. (#1763) tomtom-95andGitHub 2023-09-04 21:36:16 +02:00
• fd8e14c07a fix: unused function (#1759) 3 people 2023-09-04 20:39:50 +02:00
• c826854e48 fix: remove unused function (#1760) 3 people 2023-09-04 20:39:34 +02:00
• 2aaecc1ce4 fix: remove unused function (#1761) 3 people 2023-09-04 20:39:27 +02:00
• f863c12610 test kopt correctness (#1756) nimlgenandGitHub 2023-09-04 20:55:00 +03:00
• c6d5d45a2b Remove MemOp (#1750) George HotzandGitHub 2023-09-04 09:58:33 -07:00
• 56abe04e4b disable assembly (#1755) George HotzandGitHub 2023-09-04 09:41:20 -07:00
• b8fde6bb0f Test KOPT in CI (#1744) chenyuandGitHub 2023-09-03 14:37:20 -07:00
• ed194a1d3b zero fold (#1748) George HotzandGitHub 2023-09-03 13:48:11 -07:00
• e17b1af160 UnaryOps.NEG (#1749) George HotzandGitHub 2023-09-03 12:44:26 -07:00
• 9f1a54acee pretty kernel in cstyle (#1746) George HotzandGitHub 2023-09-03 10:21:02 -07:00
• e910e0e62c folding mul by 0 (#1743) George HotzandGitHub 2023-09-03 09:04:12 -07:00
• 3151d91f6e 3x3 winograd convs (#1675) David HouandGitHub 2023-09-03 07:29:43 -07:00
• c8025c319c Remove Token from abstractions.py (#1741) crankygrumpsterandGitHub 2023-09-02 23:56:11 -05:00
• e36148b1ce Make __getitem__ TINYer (#1661) geohotstanandGitHub 2023-09-03 11:01:01 +08:00