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

• e356807696 tinytqdm.set_description and tinytrange (#5101) chenyuandGitHub 2024-06-22 14:45:06 -04:00
• 8080298739 s/tinytqdm/tqdm (#5103) chenyuandGitHub 2024-06-22 14:18:26 -04:00
• 9f875123b6 small changes from lowerer. [run_process_replay] [no_assert] (#5102) George HotzandGitHub 2024-06-22 11:09:35 -07:00
• e468601226 update llama attention casting (#5096) chenyuandGitHub 2024-06-22 10:57:17 -04:00
• ca021229e4 fix attention to always return in the same dtype as input (#5100) chenyuandGitHub 2024-06-22 10:34:57 -04:00
• 2dcef5a0d7 hcq spec (#5081) nimlgenandGitHub 2024-06-22 15:32:12 +03:00
• 8bd6cb9511 update llama model RMSNorm casting (#5095) chenyuandGitHub 2024-06-21 23:02:04 -04:00
• 0c857ae2d6 some onnx_ops cleanups (#5094) chenyuandGitHub 2024-06-21 22:01:32 -04:00
• f4a041af16 Simplify graph_dedup [run_process_replay] (#5084) kormannandGitHub 2024-06-21 21:12:30 +02:00
• 00593d6095 clean the long lines in avg_pool2d and max_pool2d (#5091) chenyuandGitHub 2024-06-21 14:46:56 -04:00
• a971dc6218 argmax(axis=None) is argmax.flatten().argmax(0) (#5090) chenyuandGitHub 2024-06-21 14:17:10 -04:00
• 166a2b19b5 fix reduce axis of 0d tensors (#5089) chenyuandGitHub 2024-06-21 13:51:40 -04:00
• 3ff048b68c type annotate reduce axis in tensor.py (#5088) chenyuandGitHub 2024-06-21 13:06:10 -04:00
• 36b4a492a1 explicitly check getitem indices can have at most one ellipsis (#5087) chenyuandGitHub 2024-06-21 12:33:18 -04:00
• f1e758bacb graph fuzzer (#5082) nimlgenandGitHub 2024-06-21 18:47:23 +03:00
• 5717a54b28 don't use Tensor.empty in kernel opts tests (#5086) qazalandGitHub 2024-06-21 18:41:03 +03:00
• 8aa786232d docs for running process replay locally (#5083) qazalandGitHub 2024-06-21 16:55:08 +03:00
• fb1bf48cfe io_uring for copies from disk (#5035) nimlgenandGitHub 2024-06-21 11:36:51 +03:00
• b69afc67d8 tinybox docs typo geohot 2024-06-20 17:58:40 -07:00
• 6bc5e5f41c start tinybox docs geohot 2024-06-20 17:04:45 -07:00
• f6d6760f71 don't cast tuple to list before creating Tensor (#5071) chenyuandGitHub 2024-06-20 13:32:56 -04:00
• 97f1347dd9 fix check_process_replay for special characters (#5072) qazalandGitHub 2024-06-20 20:23:29 +03:00
• 6f6b3b10c9 import from uops, not linearizer (#5064) George HotzandGitHub 2024-06-20 08:08:44 -07:00
• 50700171ef minor cleanup to reshape arg handling (#5070) chenyuandGitHub 2024-06-20 10:27:27 -04:00
• f4355d0f1b check Tensor.permute input arg is a valid permutation (#5069) chenyuandGitHub 2024-06-20 10:01:28 -04:00
• 24c89a2a33 move assert_equiv_uops to helpers + use == for dtypes (#5067) qazalandGitHub 2024-06-20 16:39:34 +03:00
• e8f39fcaaa check arg to Tensor.flip can appear only once (#5068) chenyuandGitHub 2024-06-20 09:33:42 -04:00
• 55e02cdd84 generic gate folding (#5061) qazalandGitHub 2024-06-20 16:10:08 +03:00
• bdca2da2be typannos (#5059) kormannandGitHub 2024-06-20 15:02:31 +02:00
• 5f7edc7a46 minor cleanup getting output shape of _pool (#5065) chenyuandGitHub 2024-06-20 09:00:48 -04:00
• a6a5dba637 Revert "UPat for has_valid in load/store (#5052)" (#5056) qazalandGitHub 2024-06-19 20:53:36 +03:00
• ee01e464e3 use process replay as a diff creator (#4903) qazalandGitHub 2024-06-19 18:17:31 +03:00
• 99fc275c27 UPat line savings [run_process_replay] (#5053) qazalandGitHub 2024-06-19 12:43:20 +03:00
• 71194df1da UPat for has_valid in load/store [run_process_replay] (#5052) qazalandGitHub 2024-06-19 12:20:22 +03:00
• 996788358d minor change to gelu (#5048) chenyuandGitHub 2024-06-18 22:26:56 -04:00
• 4c7e316ded update pylint for ops_python (#5046) chenyuandGitHub 2024-06-18 20:15:34 -04:00
• acb715c64c fix: llama3 special tokens (#5045) wozeparrotandGitHub 2024-06-19 00:08:44 +00:00
• a8e9307e0b pylint runtime/ and shape/ (#5044) chenyuandGitHub 2024-06-18 19:48:18 -04:00
• cc2be9064f fix out of bound python list into numpy array (#5043) chenyuandGitHub 2024-06-18 18:05:21 -04:00
• 4e5add4d01 move test_tqdm to test/unit/ (#5042) chenyuandGitHub 2024-06-18 17:41:39 -04:00
• 2b2488f2e2 revert creating Tensor from a list without numpy (#5041) chenyuandGitHub 2024-06-18 17:31:22 -04:00
• e2c5054bdd update resnet.load_from_pretrained (#5040) chenyuandGitHub 2024-06-18 16:29:22 -04:00
• a3ed4176c8 use tinytqdm in active tests and examples (#5038) chenyuandGitHub 2024-06-18 16:01:19 -04:00
• fe332464d2 src->vin [run_process_replay] (#5036) kormannandGitHub 2024-06-18 21:23:49 +02:00
• f171006ded Should this symbolic test fail? (#4501) 4 people 2024-06-18 15:21:26 -04:00
• 7c3b877216 rename uop [run_process_replay] (#5031) kormannandGitHub 2024-06-18 20:34:05 +02:00
• dc942bf1f6 jit sampling functionn in test_randomness.test_multinomial (#5034) chenyuandGitHub 2024-06-18 14:21:05 -04:00
• f31ef11537 Better default hparams for large BS (#5030) Elias WahlandGitHub 2024-06-18 17:13:06 +02:00
• 8d33998e0d [run_process_replay] linearizer: fix get_grouping_dims to respect global/local max (#4855) Francis LamandGitHub 2024-06-18 06:51:27 -07:00
• 7842559952 simplification of exp2 (#5023) joeshmoe0112358andGitHub 2024-06-18 06:51:16 -07:00
• acc8f5e30e print_tree for uops (#5028) kormannandGitHub 2024-06-18 15:36:14 +02:00
• c8cd6e725c Remove BinaryOps.SUB. Replace SUB by ADD and NEG in all tests. Regenerate dataset (#4977) 3 people 2024-06-18 06:06:13 -07:00
• 620fa6e5a2 check Tensor.reshape can have at most one -1 (#5026) chenyuandGitHub 2024-06-18 08:17:12 -04:00
• 7bfa9101c0 Float in scaled dot product attention (#4985) Elias WahlandGitHub 2024-06-18 14:16:41 +02:00
• 194a168630 hcq signal scheduler (#5016) nimlgenandGitHub 2024-06-18 14:02:21 +03:00
• e9c6a36894 remove CACHELEVEL=0 in llama3 benchmark (#5025) chenyuandGitHub 2024-06-17 22:43:16 -04:00
• acaf9a490d RECIP(-0.0) should be -inf (#5024) chenyuandGitHub 2024-06-17 22:26:58 -04:00
• 66760ae558 graph display floats rounded (#5021) 3 people 2024-06-17 22:22:55 -03:00
• 03b367c014 handle float16 overflow in PYTHON (#5022) chenyuandGitHub 2024-06-17 21:12:52 -04:00
• c0139b05d8 python_alu sin(inf) is nan (#5020) chenyuandGitHub 2024-06-17 19:47:30 -04:00
• 4296507021 Tensor.sum returns in acc_dtype if specified (#5012) chenyuandGitHub 2024-06-17 16:35:52 -04:00
• 013c73c3b3 minor refactor overflow handing in python backend (#5015) chenyuandGitHub 2024-06-17 12:18:38 -04:00
• 1ad3b25461 fix einsum output str (#4998) RayandGitHub 2024-06-17 12:18:14 -04:00
• 794acefbf3 hcq update waits and signals in place (#4984) nimlgenandGitHub 2024-06-17 17:19:07 +03:00
• 603a4a0ce1 process replay contributor docs (#5010) qazalandGitHub 2024-06-17 16:38:59 +03:00
• 026c59543c allow keyword args in UOp.store [run_process_replay] (#5008) qazalandGitHub 2024-06-17 15:42:27 +03:00
• f1de8cd8cf Convert a bunch more rules [run_process_replay] (#5007) uuuvnandGitHub 2024-06-17 15:16:31 +03:00
• c52352bd9a fix yolov8 example (#5003) chenyuandGitHub 2024-06-16 20:47:29 -04:00
• 8bc0cbf67b nv tiny cleanups (#5001) nimlgenandGitHub 2024-06-17 00:43:44 +03:00
• 04feeb37e6 look for unsafe pad ops in multiview ShapeTrackers (#5002) qazalandGitHub 2024-06-17 00:28:12 +03:00
• bee8fc29ee add GPT2 half/half+beam to AMD (#5000) George HotzandGitHub 2024-06-16 14:07:14 -07:00
• 72c9b22833 sort vars in jit when building expected input args (#4990) chenyuandGitHub 2024-06-16 15:55:51 -04:00
• 71aad183fd check Program from HEAD [run_process_replay] (#4996) qazalandGitHub 2024-06-16 20:12:30 +03:00
• 2b07847f2b matmul returns in acc_dtype if specified (#4994) chenyuandGitHub 2024-06-16 12:56:15 -04:00
• 1d6f1a15e1 add lt and ge uop methods [run_process_replay] (#4995) George HotzandGitHub 2024-06-16 09:33:53 -07:00
• 1b3f27565a Boring UOps to UPat compiler [run_process_replay] (#4991) uuuvnandGitHub 2024-06-16 19:03:41 +03:00
• dac96f177e ignore indexing in the flopcounter (#4993) George HotzandGitHub 2024-06-16 08:59:55 -07:00
• 01b26756d6 Multireduce Scheduler Tests (#4972) TimmyandGitHub 2024-06-16 06:30:22 -07:00
• 5eb8001514 minor cleanup in jit (#4989) chenyuandGitHub 2024-06-15 23:43:17 -04:00
• 44dfa37c70 use threefry in stable diffusion benchmark (#4988) chenyuandGitHub 2024-06-15 20:25:29 -04:00
• 20b50d8d64 doc: manual_seed (#4987) chenyuandGitHub 2024-06-15 19:57:26 -04:00
• ce1ed374c9 more tinychat fixes (#4971) wozeparrotandGitHub 2024-06-15 23:29:39 +00:00
• 50bc14d186 re-enable test that loads torch pkl format (#4986) chenyuandGitHub 2024-06-15 14:11:30 -04:00
• ff8e9eefc3 hotfix: don't use ASSERT_COMPILE for benchmarks process replay (#4981) qazalandGitHub 2024-06-15 21:57:47 +08:00
• 92f49efd06 Trigger process replay from pull request title [run_process_replay] (#4980) uuuvnandGitHub 2024-06-15 16:21:00 +03:00
• 033fb53f9e Incomplete/buggy rule breaks process replay on #4976 (#4978) 3 people 2024-06-15 15:18:35 +03:00
• d91f0ee85b add regression test for the neg folding pattern (#4979) qazalandGitHub 2024-06-15 20:08:28 +08:00
• dfadf82e10 hcq optimize enqueue time (#4973) nimlgenandGitHub 2024-06-15 10:47:25 +03:00
• 5f7dd74655 docs: update wording for unflatten (#4974) chenyuandGitHub 2024-06-14 23:12:41 -04:00
• efbf4fca05 perf: graph_rewrite line reduction and make it a little bit faster [run_process_replay] (#4958) Cyril RoumégousandGitHub 2024-06-15 01:37:27 +02:00
• 8209cd3c55 easier llama3 + fetch subdir (#4938) wozeparrotandGitHub 2024-06-14 20:47:27 +00:00
• 64cda3c481 raise TypeError calling len() on a 0-d tensor (#4970) chenyuandGitHub 2024-06-14 16:34:27 -04:00
• 67e8df4969 remove numpy from dtype (#4969) chenyuandGitHub 2024-06-14 15:38:45 -04:00
• 62dc36d371 autogen _try_dlopen (#4949) wozeparrotandGitHub 2024-06-14 19:12:18 +00:00
• 3e297d8216 delete Linearizer.const [run_process_replay] (#4967) qazalandGitHub 2024-06-15 02:51:37 +08:00
• 118c9fe468 Tensor._fromcpu -> Tensor._fromnp (#4966) chenyuandGitHub 2024-06-14 14:33:43 -04:00
• 2a974ff257 fix: no readablestream await of, too new (#4965) wozeparrotandGitHub 2024-06-14 18:22:19 +00:00
• 9436cd4551 hcq add memory_barrier (#4963) nimlgenandGitHub 2024-06-14 21:02:55 +03:00
• dae1c8abe2 create Tensor from bytes without numpy (#4964) chenyuandGitHub 2024-06-14 13:37:27 -04:00
• 5eee974b2a construct Tensor from python list/tuple directly (#4947) chenyuandGitHub 2024-06-14 11:36:05 -04:00