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

• bd77efda2f add HWCommandQueue base class for hcq devices (#5303) nimlgenandGitHub 2024-07-11 16:19:13 +03:00
• dc3ea78560 hotfix: faster UOps.END* insert [run_process_replay] (#5377) qazalandGitHub 2024-07-11 13:20:19 +03:00
• 004366b193 context aware process replay [run_process_replay] (#5378) qazalandGitHub 2024-07-11 13:07:28 +03:00
• 45e1b9d5e3 use TC options as ContextVars [run_process_replay] (#5379) qazalandGitHub 2024-07-11 12:01:36 +03:00
• 289fd2e940 Lowerer cleanup 2 [run_process_replay] (#5376) qazalandGitHub 2024-07-11 10:56:53 +03:00
• 9ca2d96b6b delete extra check in DEFINE_ACC [run_process_replay] (#5375) qazalandGitHub 2024-07-11 10:49:03 +03:00
• 3e9f200905 KernelInfo + cleanups [run_process_replay] (#5372) George HotzandGitHub 2024-07-10 21:00:31 -07:00
• 2396ab9b33 more transcend cleanup [run_process_replay] (#5369) chenyuandGitHub 2024-07-10 23:05:03 -04:00
• 909ad72c53 remove getattr [run_process_replay] (#5370) George HotzandGitHub 2024-07-10 19:42:17 -07:00
• 0215c952c5 Move transcendental to UOp level (#5367) 3 people 2024-07-10 19:06:25 -07:00
• 64986f949c more transcend math tests in ci (#5368) chenyuandGitHub 2024-07-10 21:19:09 -04:00
• c9b3ae6bbf fix llama.py chat mode assert (#5366) wozeparrotandGitHub 2024-07-11 01:06:14 +00:00
• d13654a820 move uopgraph to file [run_process_replay] (#5364) George HotzandGitHub 2024-07-10 17:34:50 -07:00
• 320e7ed935 Approximations for SIN/LOG2/EXP2 passing all tests. (#5187) 4 people 2024-07-11 08:44:58 +09:00
• 7c0a657f08 hotfix: put process replay back geohot 2024-07-10 16:32:50 -07:00
• 7a014d5435 move ast to the kernel (#5362) George HotzandGitHub 2024-07-10 16:22:26 -07:00
• 245d83a392 more tinybox docs (#5361) wozeparrotandGitHub 2024-07-10 23:13:24 +00:00
• 6972a2569f Linearizer -> Lowerer (#4957) George HotzandGitHub 2024-07-10 15:07:42 -07:00
• 204b6169ca minimize view _reshape_mask API [run_process_replay] (#5359) chenyuandGitHub 2024-07-10 17:13:01 -04:00
• fa873df9c1 bring tinychat more inline with tinyos' version (#5358) wozeparrotandGitHub 2024-07-10 20:13:52 +00:00
• 84839f6c58 mini refactor: simpler view.unbind (#5356) Roelof van DijkandGitHub 2024-07-10 15:29:56 +02:00
• f01192b8cd tiny View.unbind cleanup [run_process_replay] (#5355) chenyuandGitHub 2024-07-09 23:37:53 -04:00
• cc80005377 clean up view _reshape_mask [run_process_replay] (#5354) chenyuandGitHub 2024-07-09 18:48:32 -04:00
• 322c37e621 use helpers.JIT in llama and gpt2 examples (#5350) chenyuandGitHub 2024-07-09 15:04:43 -04:00
• 097268fab3 Add layerwise performance bench for bert (#5349) Elias WahlandGitHub 2024-07-09 21:03:25 +02:00
• 1678199b15 add update_copy to hcq spec (#5348) nimlgenandGitHub 2024-07-09 20:44:44 +03:00
• 9504db1a57 remove the realize in _rebuild_tensor_v2 (#5347) chenyuandGitHub 2024-07-09 12:28:52 -04:00
• 1f5de80eba multi reduce Tensor.var passing verify_lazyop (#5346) qazalandGitHub 2024-07-09 17:20:17 +03:00
• 3d452195e4 [bug fix] nested commutative pattern _match [run_process_replay] [no_assert] (#5340) 3 people 2024-07-09 15:38:39 +02:00
• e815c57039 use hcq_profile in nv/amd program (#5344) nimlgenandGitHub 2024-07-09 15:56:06 +03:00
• bee96a19ff fuzz uop schedules (#5345) qazalandGitHub 2024-07-09 15:24:56 +03:00
• d5a68ae6b3 Simple abstractions3.py fix (#5343) Ian PaulandGitHub 2024-07-09 10:48:42 +00:00
• a2a9bfd2ec nv correct error messages with ptx (#5341) nimlgenandGitHub 2024-07-09 10:39:39 +03:00
• c13da83f12 tests from lowerer branch (#5339) George HotzandGitHub 2024-07-08 21:23:19 -07:00
• 4ceab5d2b1 fix PTX match rule for gated LOAD (#5338) chenyuandGitHub 2024-07-08 22:25:03 -04:00
• a80f2df1bd fix some PTX tests (#5337) chenyuandGitHub 2024-07-08 21:33:05 -04:00
• 9150a6be7a tensor metadata (#5271) wozeparrotandGitHub 2024-07-09 00:45:40 +00:00
• 7f642aa7ed minor PTX matcher cleanup [run_process_replay] (#5336) chenyuandGitHub 2024-07-08 19:19:20 -04:00
• 0f0940225a fix Tensor.all and Tensor.any for PTX (#5335) chenyuandGitHub 2024-07-08 18:15:04 -04:00
• 053c706961 refactor: expr_view on View (#5315) Roelof van DijkandGitHub 2024-07-08 20:47:34 +02:00
• 2349d837fb Fix scope order in graph toposort [run_process_replay] (#5330) kormannandGitHub 2024-07-08 20:46:15 +02:00
• 631bc974a0 raise line count limit to 8500 (#5331) chenyuandGitHub 2024-07-08 14:00:28 -04:00
• bb7746985f multireduce scheduler tests (#5141) 3 people 2024-07-08 10:28:55 -07:00
• bb2222e488 nv default for ampere & ada (#5329) nimlgenandGitHub 2024-07-08 19:01:27 +03:00
• 51d6f372e4 nv get classes based on device (#5325) nimlgenandGitHub 2024-07-08 18:25:05 +03:00
• 7d049fc20c move getting 0 and min value of a dtype to dtype.py (#5328) chenyuandGitHub 2024-07-08 10:51:56 -04:00
• b0c5c58833 nv rm_control to rmctrl type (#5327) nimlgenandGitHub 2024-07-08 17:24:33 +03:00
• 73bddc44f6 Fix fake dataloader (#5326) Elias WahlandGitHub 2024-07-08 15:07:44 +02:00
• 6856f915d6 Tensor.any and Tensor.all (#5320) chenyuandGitHub 2024-07-07 14:36:00 -04:00
• 2029cb7047 support passing None to Tensor.clip (#5319) chenyuandGitHub 2024-07-07 13:04:22 -04:00
• 296a1a36bb update Tensor.round doc and example (#5318) chenyuandGitHub 2024-07-07 12:10:39 -04:00
• c1e330f302 Tensor.int and Tensor.bool (#5317) chenyuandGitHub 2024-07-07 11:52:58 -04:00
• 778d1cdbee nv allocate local memory dynamically (#5277) nimlgenandGitHub 2024-07-07 17:34:49 +03:00
• ae10e936e7 UOps.VECTORIZE cleanups [run_process_replay] (#5314) qazalandGitHub 2024-07-07 10:49:08 +03:00
• 77b2ce9fc9 Add UOps.VECTORIZE [run_process_replay] (#5289) 4 people 2024-07-06 23:59:57 -07:00
• 2a7282c1e1 test: delete the extra cast in cstyle load [run_process_replay] [no_assert] (#5310) qazalandGitHub 2024-07-07 09:12:49 +03:00
• cededd8eb4 minor multi cleanup (#5311) chenyuandGitHub 2024-07-06 21:55:59 -04:00
• 8a99514462 generalize the uops toposort spec to ptx (#5309) qazalandGitHub 2024-07-07 00:06:30 +03:00
• ca0ef1700b use precise::sin in metal (#5307) chenyuandGitHub 2024-07-06 12:47:27 -04:00
• 5c2ca7bad4 remove UOps.SPECIAL rendering from llvm (#5306) qazalandGitHub 2024-07-06 19:28:47 +03:00
• 356e5d2e54 touchup multi dtype in elementwise (#5305) chenyuandGitHub 2024-07-06 11:54:12 -04:00
• 7ddda9f9f1 hotfix: cache seen graphs in fusion (#5302) qazalandGitHub 2024-07-06 14:13:58 +03:00
• 11dfb19b20 track seen graphs in recursive group (#5301) qazalandGitHub 2024-07-06 12:39:31 +03:00
• d813617742 prescheduling refactor (#5300) qazalandGitHub 2024-07-06 12:04:03 +03:00
• c1e166c08a fix dtype mismatch for bool ops in multi (#5299) qazalandGitHub 2024-07-06 11:36:40 +03:00
• fc03fc025e enable sin on METAL in test_dtype_alu (#5298) chenyuandGitHub 2024-07-05 14:52:09 -04:00
• b369e75ed0 refactor schedule creation (#5297) qazalandGitHub 2024-07-05 21:14:38 +03:00
• 5292d37db6 LoadOps.VIEW in the scheduler spec (#5296) qazalandGitHub 2024-07-05 19:43:50 +03:00
• 1ab7a4cff0 Handling Multiple UnaryOps.BITCAST in Function for Proper Kernel Fusion [run_process_replay] (#5172) 3 people 2024-07-06 01:16:44 +09:00
• 43c3f73fbc handcode_bert_opt.py (#5295) chenyuandGitHub 2024-07-05 11:01:20 -04:00
• d7835a705c hotfix: fix metal with vars (#5294) nimlgenandGitHub 2024-07-05 16:53:40 +03:00
• 8a548b0b6e metal support offset (#5293) nimlgenandGitHub 2024-07-05 16:13:05 +03:00
• 1cefbb33ab uop graph tests + type_verify cleanup (#5292) qazalandGitHub 2024-07-05 13:00:01 +03:00
• 341c4a29d1 hotfix: use dtype.scalar() for rendering cast [run_process_replay] [no_assert] (#5290) qazalandGitHub 2024-07-05 11:29:35 +03:00
• 87d27c45ec minor _broadcast cleanup (#5286) chenyuandGitHub 2024-07-04 14:25:24 -04:00
• 8c03816ae9 fix README example (#5284) SnakeOnexandGitHub 2024-07-04 17:15:07 +02:00
• 2778b6046c new memory scheduler (#5278) nimlgenandGitHub 2024-07-04 18:06:04 +03:00
• 84b3e3bb6f hcq exec no embedded signal (#5142) nimlgenandGitHub 2024-07-04 13:29:21 +03:00
• 0c3a35e5c2 Stable Diffusion v2 Inference (#5283) Tobias FischerandGitHub 2024-07-03 22:47:10 -04:00
• e5ba385f03 remove first contiguous in multi from_sharded (#5121) chenyuandGitHub 2024-07-03 19:42:56 -04:00
• f1ff65e763 remove "no-nans-fp-math"="true" for LLVM (#5282) chenyuandGitHub 2024-07-03 17:52:50 -04:00
• 3929a9dc94 fix UOp.cmp_tuple for ALU (#5280) chenyuandGitHub 2024-07-03 14:59:05 -04:00
• a9d6a6c339 verify_lazyop with multi reduce (#5276) qazalandGitHub 2024-07-03 20:15:42 +03:00
• 16e3b8b013 uops work from lowerer [run_process_replay] (#5279) George HotzandGitHub 2024-07-03 09:40:00 -07:00
• 622b7bd556 simpler TinyJit inside TinyJit detection (#5219) chenyuandGitHub 2024-07-03 12:28:53 -04:00
• 04ef0fd328 fix: message when applegpu tools missiong (#5236) gipandGitHub 2024-07-03 09:07:09 -07:00
• d3e244d8b7 prev speed improvements (#5252) 3 people 2024-07-03 12:06:01 -04:00
• 21d41f06a2 nv follows HCQCompatAllocRes protocol (#5275) nimlgenandGitHub 2024-07-03 11:34:10 +03:00
• d3e4e21759 add return type for HCQCompatAllocator _alloc (#5267) 3 people 2024-07-03 10:25:44 +03:00
• 191463a919 add timing to SDXL (#5273) chenyuandGitHub 2024-07-02 23:29:54 -04:00
• b2c3a28a5e nn.RMSNorm (#5272) chenyuandGitHub 2024-07-02 21:39:01 -04:00
• 9a2a82a77f test stable diffusion unet in ci (#5268) chenyuandGitHub 2024-07-02 21:37:52 -04:00
• ce52b10f6f add a flag DISABLE_LOOP_COLLAPSE (#5270) chenyuandGitHub 2024-07-02 20:01:11 -04:00
• e53b164e1a small changes from lowerer (#5266) George HotzandGitHub 2024-07-02 15:03:54 -07:00
• 7be776f9af add _alloc_signal/_free_signal to hcq (#5264) nimlgenandGitHub 2024-07-02 23:35:39 +03:00
• 9a25ee0b9a pixed unet call params (#5262) Tobias FischerandGitHub 2024-07-02 12:40:27 -04:00
• 59bc837ad1 refactor gated load rendering [run_process_replay] (#5259) qazalandGitHub 2024-07-02 15:13:10 +03:00
• e050603b4b nv close fds after mapping (#5246) nimlgenandGitHub 2024-07-02 13:57:46 +03:00
• d3cfb6c2e3 refactor UOps.LOAD barrier [run_process_replay] (#5258) qazalandGitHub 2024-07-02 13:48:47 +03:00
• a1044e6063 iterate over scoped uops once [run_process_replay] (#5255) qazalandGitHub 2024-07-02 09:21:09 +03:00