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

• eb77488f85 update llama3 70B to use R1 (#8733) chenyuandGitHub 2025-01-23 19:06:05 -05:00
• 3e987fc856 add device print with -m tinygrad.device [pr] (#8729) George HotzandGitHub 2025-01-23 12:46:27 -08:00
• 04846b91aa reorder and categorize onnx_ops (#8731) geohotstanandGitHub 2025-01-24 02:18:54 +08:00
• 8e5bd0cd7a fix buffer init and skip test_swizzle_failure_permute [pr] (#8732) qazalandGitHub 2025-01-23 10:21:38 -05:00
• e4512baea4 am: cleanup mm (#8730) nimlgenandGitHub 2025-01-23 15:49:37 +03:00
• 07ec99001a keep VIEW in big_sink + copy of buffer view spec [pr] (#8727) qazalandGitHub 2025-01-23 04:29:30 -05:00
• 6cb74bb630 fix using clone with shrink [pr] (#8724) qazalandGitHub 2025-01-23 01:28:07 -05:00
• af65331b76 update beam params for bert green [pr] (#8726) chenyuandGitHub 2025-01-22 22:00:05 -05:00
• 907dfa0e82 image buffer realization spec [pr] (#8420) qazalandGitHub 2025-01-22 13:25:22 -05:00
• 49b914ee69 simpler bert acc [pr] (#8714) chenyuandGitHub 2025-01-22 10:32:19 -05:00
• 93fb50ce77 allreduce: add flags (#8713) nimlgenandGitHub 2025-01-22 17:44:31 +03:00
• 891436853d remove buffer size check in schedule item [pr] (#8712) qazalandGitHub 2025-01-22 06:36:30 -05:00
• 2dae467b75 scheduler + process_replay import cleanup (#8711) qazalandGitHub 2025-01-22 05:44:07 -05:00
• e3d1464ba4 move assign preload out of schedule item [pr] (#8710) qazalandGitHub 2025-01-22 05:43:57 -05:00
• 9a9079118e envvar BERT_LAYERS [pr] (#8709) chenyuandGitHub 2025-01-21 22:49:19 -05:00
• 9f6d545a16 bert log global_norm in training step [pr] (#8708) chenyuandGitHub 2025-01-21 20:36:27 -05:00
• c5e46c5eee am: recover from any boot interrupt (#8703) nimlgenandGitHub 2025-01-21 22:22:23 +03:00
• 1e283c33d3 remove realize in bert model init [pr] (#8707) chenyuandGitHub 2025-01-21 14:11:03 -05:00
• 018edd934b don't use view in copy [pr] (#8704) George HotzandGitHub 2025-01-21 09:57:47 -08:00
• d6bf1feaab remove the "no copy" line from copy_to_device (#8702) qazalandGitHub 2025-01-21 10:09:33 -05:00
• 3628f89929 fix deallocate for subbuffers (#8701) nimlgenandGitHub 2025-01-21 16:34:19 +03:00
• 6733a3a96b am: fix typo (#8700) nimlgenandGitHub 2025-01-21 14:35:15 +03:00
• f0d424ecdf Tensor UOps can become a buffer or const after scheduling (#8698) qazalandGitHub 2025-01-21 05:33:19 -05:00
• e2008c98c3 allow symbolic shape in tensor const parents [pr] (#8699) qazalandGitHub 2025-01-21 05:01:25 -05:00
• 2b239db5d2 temp() with usernames (#8697) nimlgenandGitHub 2025-01-21 12:26:43 +03:00
• 66ac0087e8 more high level contiguous tests + scheduler deletions [pr] (#8695) qazalandGitHub 2025-01-20 18:52:58 -05:00
• 08eb1f1f56 simplify tensors before scheduling [pr] (#8580) 3 people 2025-01-20 16:42:42 -05:00
• 02ad450e22 add failing assert for gradient realization [pr] (#8692) qazalandGitHub 2025-01-20 15:50:09 -05:00
• b14c9848cc small changes to make the tensor_map_simple diff cleaner [pr] (#8691) qazalandGitHub 2025-01-20 15:25:59 -05:00
• 1a15c0e89d Move define_acc down an unrolled add chain (#8404) 3 people 2025-01-20 20:56:27 +01:00
• dd82b4c913 make onnx runner a class (#8647) 3 people 2025-01-21 02:11:05 +08:00
• 46a8c5e1e5 delete forced_realize (#8615) 4 people 2025-01-20 09:40:36 -08:00
• 679b1ad058 move softmax upcast to after subtracting max (#8684) chenyuandGitHub 2025-01-20 12:16:32 -05:00
• 08ca871d77 am: remove pm block (#8688) nimlgenandGitHub 2025-01-20 18:05:22 +03:00
• 9d3c40601f am: fast memory manager (#8654) nimlgenandGitHub 2025-01-20 16:58:22 +03:00
• 9e55495b4d fold double contiguous [pr] (#8687) qazalandGitHub 2025-01-20 07:38:33 -05:00
• ed63ff2372 Remove contiguous on buffer (#8676) qazalandGitHub 2025-01-20 06:48:33 -05:00
• 3499a2c72d start moving image things to rewrite rules (#8678) qazalandGitHub 2025-01-20 06:34:29 -05:00
• b1847d561f smaller do_realize and some cleanups [pr] (#8685) qazalandGitHub 2025-01-20 05:47:01 -05:00
• 689bf68cfc remove GroupOp.Meta [pr] (#8686) qazalandGitHub 2025-01-20 05:24:19 -05:00
• 4198bce150 _apply_map_to_tensors [pr] (#8683) George HotzandGitHub 2025-01-19 17:56:04 -08:00
• 98d01a059d rename uopgraph to rewriter [pr] (#8682) George HotzandGitHub 2025-01-19 17:03:12 -08:00
• f532c78889 minor space hotfix (#8679) Ignacio SicaandGitHub 2025-01-19 22:00:24 -03:00
• 2d0842386d fix parse_valid for float uop (#8681) chenyuandGitHub 2025-01-19 18:15:49 -05:00
• 168c16646a change create_schedule_with_vars api to big_sink [pr] (#8677) George HotzandGitHub 2025-01-19 13:30:26 -08:00
• beba490ba8 update mask in scaled_dot_product_attention (#8674) chenyuandGitHub 2025-01-19 15:19:23 -05:00
• 5842ee56c6 raise if attn_mask is set when is_causal=True in sdpa [pr] (#8675) chenyuandGitHub 2025-01-19 12:55:04 -05:00
• 2faf8774fe replace DEVICE of CONST after copy folding (#8673) qazalandGitHub 2025-01-19 11:33:39 -05:00
• d957a4f108 add tests for div buffer collapsing in the scheduler [pr] (#8671) qazalandGitHub 2025-01-18 14:15:29 -05:00
• bd0fb14d70 hotfix: add ctx to VIZ rewrite (#8667) qazalandGitHub 2025-01-18 07:58:47 -05:00
• 0ef85b52e6 init folding changes from the tensor_map branch [pr] (#8666) qazalandGitHub 2025-01-18 07:15:52 -05:00
• 5267a411e7 remove movementops in viz graph rewrite [pr] (#8665) qazalandGitHub 2025-01-18 06:34:12 -05:00
• b49a04145e fix for int plus minor cleanup (#8650) ignaciosicaandGitHub 2025-01-18 00:30:39 -03:00
• c49e0fca60 GlobalCounters.reset() in sdxl step [pr] (#8664) chenyuandGitHub 2025-01-17 21:10:28 -05:00
• d2234e308a tf32 tc for nv and ptx (#8635) 3 people 2025-01-17 22:43:57 -03:00
• 5afb0a4a81 metal: fix transfer profiling (#8659) nimlgenandGitHub 2025-01-17 23:47:01 +03:00
• 0d7bd4f389 empty graph rewrite to VIZ tensor graph [pr] (#8658) George HotzandGitHub 2025-01-17 11:29:33 -08:00
• 8609b880bd hotfix: test_backward_sum geohot 2025-01-17 10:24:10 -08:00
• f8cc971c3b raise RuntimeError for uneven shards in Tensor.shard [pr] (#8656) chenyuandGitHub 2025-01-17 12:48:39 -05:00
• 3506a7585f upcast overflowed idx to int64 [pr] (#8268) 3 people 2025-01-18 00:52:31 +08:00
• 23f0ff0ed8 add bitcast to multi [pr] (#8652) qazalandGitHub 2025-01-17 03:17:19 -05:00
• 2b7db9b45d delete unused cast/bitcast lines from ops.py [pr] (#8651) qazalandGitHub 2025-01-17 03:04:18 -05:00
• 4f0d1b4759 Disable graphs by default if using an intel macbook (#8648) (#8649) Mike AshcroftandGitHub 2025-01-16 21:24:56 -05:00
• 0289fbb1c2 limit real_size to the size of first View of ShapeTracker (#8628) 3 people 2025-01-16 21:27:39 +00:00
• f91ca508cf am: bind for sdma (#8633) nimlgenandGitHub 2025-01-16 15:22:27 +03:00
• f671da6755 ci: add AM start time to benchmark (#8637) nimlgenandGitHub 2025-01-16 14:47:36 +03:00
• 81a84aa85a remove is_unrealized_unmasked_const [pr] (#8644) qazalandGitHub 2025-01-16 05:27:47 -05:00
• 00e5979897 Use full soname for libgcc_s in CPUProgram (#8642) uuuvnandGitHub 2025-01-16 11:56:52 +02:00
• 611208cd8a Revert "Revert "move subbuffer to a rewrite rule in the scheduler (#8639)" (…" (#8643) qazalandGitHub 2025-01-16 04:30:11 -05:00
• 82ef956cb8 Revert "move subbuffer to a rewrite rule in the scheduler (#8639)" (#8641) qazalandGitHub 2025-01-16 03:29:07 -05:00
• d5c90da286 move subbuffer to a rewrite rule in the scheduler (#8639) qazalandGitHub 2025-01-15 20:14:28 -05:00
• b3efeeb717 docs: start am docs (#8638) nimlgenandGitHub 2025-01-16 00:22:35 +03:00
• 7ecced7f6d LLVM JIT prereqs (#8634) uuuvnandGitHub 2025-01-15 19:47:08 +02:00
• a1f70ce7d0 only use BUFFER_VIEW in disk [pr] (#8629) qazalandGitHub 2025-01-15 12:34:15 -05:00
• bae20e5043 Generic PTX wmma rendering [pr] (#8632) ignaciosicaandGitHub 2025-01-15 14:31:48 -03:00
• 6193e279d4 isolate simple failing test for subbuffer on CONST [pr] (#8630) qazalandGitHub 2025-01-15 05:45:03 -05:00
• e1f7c90459 gradient is a set [pr] (#8626) George HotzandGitHub 2025-01-14 20:48:23 -08:00
• 7fb1c7af61 minor multi cleanups [pr] (#8625) chenyuandGitHub 2025-01-14 22:25:23 -05:00
• 504ad08e73 hotfix: add test_example_matmul_same geohot 2025-01-14 19:03:17 -08:00
• f29d6f54b8 support multilb gradient [pr] (#8624) George HotzandGitHub 2025-01-14 18:33:33 -08:00
• 4ee3243c93 JITBEAM=2 for LLaMA-3 8B on 4 GPUs [pr] (#8623) chenyuandGitHub 2025-01-14 19:52:38 -05:00
• 7860a80801 simpler MultiLazyBuffer alu [pr] (#8622) chenyuandGitHub 2025-01-14 19:19:13 -05:00
• 930728c069 bert BS 72->66 [pr] (#8621) chenyuandGitHub 2025-01-14 18:41:41 -05:00
• 0790d8059f remove MultiLazyBuffer.from_sharded [pr] (#8620) chenyuandGitHub 2025-01-14 18:00:49 -05:00
• c85737c200 assert to prepare for grad uop [pr] (#8280) 3 people 2025-01-14 13:26:56 -08:00
• fdd46c9f28 delete view instant rule (#8616) 4 people 2025-01-14 13:15:13 -08:00
• dddd4e5f9f hotfix: remove duplicate TestTensorMutates [pr] (#8619) qazalandGitHub 2025-01-14 16:03:17 -05:00
• c5782e85d2 tlsf: optimize alloc (#8608) nimlgenandGitHub 2025-01-14 23:48:07 +03:00
• bfbe81df71 remove cast before view (#8613) 3 people 2025-01-14 12:04:58 -08:00
• 393eec3201 raise RuntimeError for uneven shard [pr] (#8593) chenyuandGitHub 2025-01-14 14:51:48 -05:00
• d5a646d492 CUDA Turing TC (#8597) 3 people 2025-01-14 15:35:14 -03:00
• cbfd51f5a5 make MultiLazyBuffer.bounds a property [pr] (#8614) chenyuandGitHub 2025-01-14 13:25:54 -05:00
• 52e7003414 Revert "make kits19 dataset samples have small sizes (#8591)" (#8610) chenyuandGitHub 2025-01-14 12:24:27 -05:00
• 76a03e950a make kits19 dataset samples have small sizes (#8591) Francis LataandGitHub 2025-01-14 11:27:45 -05:00
• 4057b98f7f rename i and j into k and row/col (#8607) ignaciosicaandGitHub 2025-01-14 13:27:05 -03:00
• 1ff6862a3d ci: sleep a bit to let the driver unload the prev pid (#8605) nimlgenandGitHub 2025-01-14 15:55:23 +03:00
• 97ec564b03 noop changes from the block_assign branch [pr] (#8606) qazalandGitHub 2025-01-14 07:47:17 -05:00
• 5aab2806f0 rename to test_tensor_uop + use upats for asserting [pr] (#8604) qazalandGitHub 2025-01-14 05:09:56 -05:00
• 863abc7140 scheduling graph_rewrite prereqs for BLOCK in ASSIGN (#8598) qazalandGitHub 2025-01-14 03:01:59 -05:00
• 05e54f00d3 remove bounds from MultiLazyBuffer.from_sharded [pr] (#8603) chenyuandGitHub 2025-01-13 23:40:05 -05:00