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

• 25ec40ca93 cleanup dtype of tensor creation from list (#4566) chenyuandGitHub 2024-05-13 02:47:41 -04:00
• 4e1135a0bc assign buffer read/write tests (#4565) qazalandGitHub 2024-05-13 14:43:36 +08:00
• b660f60125 all uops are now cachable (#4564) George HotzandGitHub 2024-05-12 22:34:35 -07:00
• 02327b8adf simple stuff from new_uops branch (#4563) George HotzandGitHub 2024-05-12 22:18:05 -07:00
• f53a23d21e Test for optim assertion (#4558) 3 people 2024-05-12 23:21:28 +02:00
• d7670f8141 quantized llama multilazybuffer fix (#4557) wozeparrotandGitHub 2024-05-12 14:19:21 -07:00
• bcee4743ce fix error message (#4556) 3 people 2024-05-12 21:35:51 +02:00
• 01a0c1a948 slightly faster nf4 llama (#4542) chenyuandGitHub 2024-05-12 14:24:42 -04:00
• 4c232dc0ae refactor LoadOps scheduling (#4553) qazalandGitHub 2024-05-12 17:59:24 +08:00
• 3da152f0fe scheduler docs 2 (#4551) qazalandGitHub 2024-05-12 17:15:39 +08:00
• e07c7668b3 nf4 llama (#4540) wozeparrotandGitHub 2024-05-11 22:22:34 -07:00
• 7a26bdac65 move scheduleitem to schedule.py (#4541) George HotzandGitHub 2024-05-11 21:13:04 -07:00
• 508e8a6666 add cpu objdump to LLVM/CLANG (#4537) George HotzandGitHub 2024-05-11 14:28:44 -07:00
• bed70b130c mlperf bert getenv-able EVAL_STEP_FREQ (#4534) chenyuandGitHub 2024-05-11 14:36:56 -04:00
• 328b083e66 lil profiling script geohot 2024-05-11 11:02:44 -07:00
• da10cf0be1 extra/threefry.py for mem usage (#4533) chenyuandGitHub 2024-05-11 13:46:44 -04:00
• 8a0fb3d765 delete old extra/autopad.py (#4532) chenyuandGitHub 2024-05-11 13:06:10 -04:00
• 04a4980a51 touchup bert script (#4531) chenyuandGitHub 2024-05-11 13:02:02 -04:00
• 4871476a1e move copy kernel to out of schedule ordering (#4530) qazalandGitHub 2024-05-11 19:44:44 +08:00
• 2fb564c125 multi reduce linearizer tests start (#4529) qazalandGitHub 2024-05-11 19:06:40 +08:00
• 3cba22920f test_linearizer_correctness (#4458) qazalandGitHub 2024-05-11 18:02:08 +08:00
• b3d9fd48d0 infra for testing linearizer correctness (#4528) qazalandGitHub 2024-05-11 17:10:33 +08:00
• 2f970a4fc2 all realize 2 (#4527) George HotzandGitHub 2024-05-10 22:43:09 -07:00
• d2c347fc74 faster gather for bert (#4526) wozeparrotandGitHub 2024-05-10 22:28:48 -07:00
• 922e6e056a hotfix: fix docs geohot 2024-05-10 21:51:35 -07:00
• 347a3acb37 add renderer class (#4524) George HotzandGitHub 2024-05-10 21:40:02 -07:00
• b00b6b16f0 fix TRAIN_BEAM and Tensor.training for mlperf bert (#4525) chenyuandGitHub 2024-05-11 00:18:36 -04:00
• 7fab8c9e17 add symbolic mean test cases in test_symbolic_ops and test_symbolic_jit (#4523) chenyuandGitHub 2024-05-10 23:19:55 -04:00
• 827058f030 update tests get_runner (#4522) George HotzandGitHub 2024-05-10 20:09:22 -07:00
• a0448ff595 use copy kernel in schedule (#4520) George HotzandGitHub 2024-05-10 15:30:33 -07:00
• b15e2309bd verbose error message in getitem (#4519) chenyuandGitHub 2024-05-10 17:25:41 -04:00
• d438d5698d bring buffer back to device (#4517) George HotzandGitHub 2024-05-10 11:22:31 -07:00
• a2b707a3eb scheduler comments 1 (#4515) qazalandGitHub 2024-05-11 01:44:28 +08:00
• 4eef1ee9bf move renderer into options (#4514) George HotzandGitHub 2024-05-10 10:01:51 -07:00
• 7c630a9a53 hotfix: fix llama spacing + fix hcq geohot 2024-05-10 15:10:13 +00:00
• 58e7256ce9 restore hcq graph (#4513) George HotzandGitHub 2024-05-10 07:45:05 -07:00
• 06c1e7498e Revert "hcq graph (#4380)" (#4512) George HotzandGitHub 2024-05-10 07:18:09 -07:00
• 84a2e2b8c1 hcq graph (#4380) nimlgenandGitHub 2024-05-10 17:15:12 +03:00
• 2b7ab60584 dfs fusion (#4491) qazalandGitHub 2024-05-10 22:00:48 +08:00
• bd8bb82555 move fusion out of child iteration (#4509) qazalandGitHub 2024-05-10 17:03:32 +08:00
• ff216a383a refactor fused children (#4508) qazalandGitHub 2024-05-10 16:49:23 +08:00
• b399d98e41 fix resnet eval (#4507) chenyuandGitHub 2024-05-10 00:49:00 -04:00
• a602dc67d3 feat: more mlperf fixes (#4505) wozeparrotandGitHub 2024-05-09 20:50:20 -07:00
• 0e8aa0e288 use fake data in beam searching resnet (#4504) chenyuandGitHub 2024-05-09 23:43:50 -04:00
• 5bfc33948a hotfix: only run optimize_local_size once geohot 2024-05-09 20:01:53 -07:00
• 29daea4e60 fix: core count and os (#4503) wozeparrotandGitHub 2024-05-09 19:55:07 -07:00
• 89e119bc58 move Allocator to buffer.py (#4502) George HotzandGitHub 2024-05-09 19:45:56 -07:00
• 1e843d495e cleaning up search with Program (#4500) George HotzandGitHub 2024-05-09 19:01:53 -07:00
• d3dc332c2e Tensor.logsumexp (#4442) chenyuandGitHub 2024-05-09 20:49:06 -04:00
• 78b298aa2a move 0d tensor reduce axis check to _reduce (#4499) chenyuandGitHub 2024-05-09 20:29:55 -04:00
• c9e84ed0da refactor to Program class (#4476) George HotzandGitHub 2024-05-09 17:29:07 -07:00
• 5de4a46f10 re-enable gpt2 half/beam mac benchmark (#4496) chenyuandGitHub 2024-05-09 19:15:32 -04:00
• a2e2ba380c nv tune shmem size (#4495) nimlgenandGitHub 2024-05-10 00:35:01 +03:00
• ef93e41a15 resnet mlperf systems add tinygrad commit and python / runtime versions (#4494) chenyuandGitHub 2024-05-09 16:04:15 -04:00
• b5afdfbc5b first draft resnet mlperf readme (#4493) chenyuandGitHub 2024-05-09 15:51:44 -04:00
• 047c7f3e5b polish resnet mlperf logging (#4490) chenyuandGitHub 2024-05-09 13:04:24 -04:00
• d78e159aa3 resnet logging move RUN_START to start of the script (#4488) chenyuandGitHub 2024-05-09 12:32:32 -04:00
• 1bcb58479d resnet setup power cap red box gpu to 350W (#4484) chenyuandGitHub 2024-05-08 23:32:41 -04:00
• 0ed755bcf5 resnet use EVAL_BS=192 (#4482) chenyuandGitHub 2024-05-08 22:29:27 -04:00
• 1f6bf9d2f7 real diskcache_clear in model_train resnet (#4445) chenyuandGitHub 2024-05-08 19:06:09 -04:00
• 1b4645bea6 hotfix resnet move init_start to start of the script (#4481) chenyuandGitHub 2024-05-08 19:03:52 -04:00
• a347ae94d6 feat: remove wandb (#4480) wozeparrotandGitHub 2024-05-08 15:31:16 -07:00
• 00c309dfe2 trigger tc in remu (#4479) qazalandGitHub 2024-05-09 04:23:46 +08:00
• e14d5b6fd7 nv fix oob qmd ptr (#4478) nimlgenandGitHub 2024-05-08 23:11:04 +03:00
• db7e15c46f hotfix resnet only log epoch start with RUNMLPERF (#4477) chenyuandGitHub 2024-05-08 15:14:41 -04:00
• 062c6dd65d mlperf logging, truncate dir in logs and log seed (#4475) chenyuandGitHub 2024-05-08 12:54:02 -04:00
• b62a65b617 redo faster sparse_categorical_crossentropy (#4461) chenyuandGitHub 2024-05-08 11:21:43 -04:00
• e87460c7e2 bump version (#4474) Elias WahlandGitHub 2024-05-08 16:48:42 +02:00
• 4eb6aef73c Speed up graph rewrite (#4473) Szymon OżógandGitHub 2024-05-08 16:15:15 +02:00
• cc33947fa5 Update links in new docs (#4363) Nicklas BomanandGitHub 2024-05-08 15:13:00 +02:00
• 36a1f38049 lazy folding: mul -1 is neg, and neg neg is noop (#4472) chenyuandGitHub 2024-05-08 01:52:22 -04:00
• c508eb7425 revert the removal of CAST_BEFORE_VIEW (#4471) chenyuandGitHub 2024-05-08 00:14:29 -04:00
• 5dbab7fae6 bring thneed back (#4467) George HotzandGitHub 2024-05-07 20:55:03 -07:00
• 7eb035e7c5 stronger test case for half mean overflow (#4470) chenyuandGitHub 2024-05-07 22:40:09 -04:00
• ca7300c783 fix half mean and its backward (#4469) chenyuandGitHub 2024-05-07 21:46:41 -04:00
• 7da1b41f38 fuzz_linearizer: add FUZZ_REQUIRE_TC option to require TC in opts (#4468) Francis LamandGitHub 2024-05-07 14:14:21 -07:00
• 46a793111b test for LazyBuffer._view when mask out and degrade into const (#4465) chenyuandGitHub 2024-05-07 12:56:23 -04:00
• a1d350a810 nv timeline semaphores (#4464) nimlgenandGitHub 2024-05-07 17:31:19 +03:00
• e3bb85fd0e amd timeline semaphores (#4416) nimlgenandGitHub 2024-05-07 11:17:32 +03:00
• 17faae091b optimizer shouldn't be run without training (#4460) George HotzandGitHub 2024-05-06 15:34:12 -07:00
• 35dfbc6354 rand_for_dtype helper (#4459) qazalandGitHub 2024-05-07 05:03:42 +08:00
• a3140c9767 nv boost subdevice (#4456) nimlgenandGitHub 2024-05-06 23:05:20 +03:00
• 47750e65fd kernel: un-reverse the order of the local indices (#4454) Francis LamandGitHub 2024-05-06 12:21:27 -07:00
• 5e036cd0b3 test unary and more reduces in test_flopcounter (#4455) chenyuandGitHub 2024-05-06 15:15:16 -04:00
• d0b8862dea fix out of resource kernels on nv (#4450) nimlgenandGitHub 2024-05-06 19:24:20 +03:00
• f4e49a7c1a resnet 50 opt: correct loop + LARS (#4449) George HotzandGitHub 2024-05-06 08:01:26 -07:00
• 292ce64ad7 move acc_dt out of lazy (#4382) chenyuandGitHub 2024-05-06 10:41:25 -04:00
• 113c2f00b9 amd doorbell size is 64bits (#4448) nimlgenandGitHub 2024-05-06 16:59:59 +03:00
• fc995d4446 add backward to handcode_resnet50_opt geohot 2024-05-06 06:38:24 -07:00
• 6dbe5585b0 batchnorm + conv backward in test_schedule (#4420) qazalandGitHub 2024-05-06 21:40:17 +08:00
• 3f3c973022 Multiple Reduce Kernels - kernel properly orders reduceops (#4418) 3 people 2024-05-06 06:54:44 -04:00
• 603d3a351b feat: allow keeping multiple cookies (#4440) wozeparrotandGitHub 2024-05-05 19:26:48 -07:00
• afe020710d disable PADTO on upcasted axis (#4444) chenyuandGitHub 2024-05-05 21:52:03 -04:00
• 709410071c mlperf/resnet: updated BEAM params to increase performance (#4443) Francis LamandGitHub 2024-05-05 18:49:46 -07:00
• c8595a9655 update sops.gz, fix tests and add new linearizer test (#4437) Francis LamandGitHub 2024-05-05 14:31:25 -07:00
• 9ad3d0520a hotfix: npy is also ok (#4439) wozeparrotandGitHub 2024-05-05 13:48:54 -07:00
• d0eb1540d5 helpers.diskcache_clear (#4436) chenyuandGitHub 2024-05-05 14:19:01 -04:00
• 595a6e3069 test_fold_conv_relu_backward test geohot 2024-05-05 11:13:43 -07:00
• cc16f644d0 hotfix: remove FAKE buffer from graph geohot 2024-05-05 10:51:52 -07:00
• 760776c59d merge EfficientNet to C with clang job (#4426) qazalandGitHub 2024-05-06 01:33:12 +08:00