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

• a77ee72d11 clean up reshape size check [pr] (#8067) chenyuandGitHub 2024-12-06 07:51:19 -05:00
• 074a67a6eb combine get inputs and type_parse function in onnx (#8069) geohotstanandGitHub 2024-12-06 20:42:35 +08:00
• c0240855b9 qcom has not transfer (#8075) nimlgenandGitHub 2024-12-06 14:45:01 +03:00
• ce72fe1411 u32 to f16 in tinygrad (#8074) Ahmed HarmoucheandGitHub 2024-12-06 12:00:13 +01:00
• e37bff6c19 fix bug in jit prune with copy [pr] (#8073) George HotzandGitHub 2024-12-06 18:38:23 +08:00
• aae8557ada test copy inside jit [pr] (#8072) George HotzandGitHub 2024-12-06 17:51:50 +08:00
• e2fe7f0d2f hotfix: actually fix pylint, it's a python 3.10 issue geohot 2024-12-06 13:53:46 +08:00
• b28d660172 update self_tokenize, fix pylint maybe geohot 2024-12-06 13:49:41 +08:00
• 344fd4845c example: self_tokenize. someday tinygrad will be recursively self improving geohot 2024-12-06 13:35:02 +08:00
• 3c5d5f9414 mypy==1.13.0 (#7990) 3 people 2024-12-06 05:09:14 +01:00
• 65b6696f3b refactor safe_load (#8035) leopfandGitHub 2024-12-06 05:08:21 +01:00
• e7d5fe4a32 improve idiv _min_max (#8066) chenyuandGitHub 2024-12-05 23:02:16 -05:00
• 13b954f22c unify expand conditions [pr] (#8065) chenyuandGitHub 2024-12-05 21:40:14 -05:00
• aefdff4ef5 reshape mask cleanups [pr] (#8064) chenyuandGitHub 2024-12-05 20:20:43 -05:00
• 05dba6e4ee minor to_indexed_uops cleanup [pr] (#8063) chenyuandGitHub 2024-12-05 17:15:03 -05:00
• b2dd703592 fix typing of UOp.range [pr] (#8062) chenyuandGitHub 2024-12-05 14:56:34 -05:00
• 49c6dab74b Add pattern for div mod recombine with gcd (#8061) 3 people 2024-12-05 19:16:58 +01:00
• 707e9a9c8e add _one_hot_along_dim helper for Tensor.arange masking (#8039) 3 people 2024-12-06 01:43:00 +08:00
• 3c5983473a combine parentless reduce rule [pr] (#8059) chenyuandGitHub 2024-12-05 11:28:35 -05:00
• 87594a8153 simpler dtypes.max for int [pr] (#8058) chenyuandGitHub 2024-12-05 10:31:41 -05:00
• 66b8242375 Simple onnx.py clean ups (#8054) 3 people 2024-12-05 23:31:26 +08:00
• 5c6ed5dba6 lower test_conv_3x3_256_32_32_256_256 expectation (#8060) chenyuandGitHub 2024-12-05 10:30:56 -05:00
• c6f5bb03fa YoloV8 WebGPU fixes (#8057) Ahmed HarmoucheandGitHub 2024-12-05 16:23:45 +01:00
• fc9ae53a05 Progress with pack/unpack wgsl built-ins wgpu-f16-packed Ahmed Harmouche 2024-12-05 15:55:09 +01:00
• 4669e467ef f16 comp/decomp WIP Ahmed Harmouche 2024-12-05 14:14:23 +01:00
• 78c01a5c2b amd general _gpu_alloc (#8056) nimlgenandGitHub 2024-12-05 15:50:23 +03:00
• 8071600897 nv one _gpu_alloc (#8055) nimlgenandGitHub 2024-12-05 15:22:03 +03:00
• ff9a89f714 Proper dtypes for input/output of exported WebGPU model (#8053) Ahmed HarmoucheandGitHub 2024-12-05 10:38:05 +01:00
• 435a51e10c reduce folding simple tests [pr] (#8040) qazalandGitHub 2024-12-04 23:22:45 -05:00
• 20878be2af lower test_gemv_4096_16384 expectations geohot 2024-12-05 12:07:38 +08:00
• 83aecbdc70 do gpuocelot copy manually [pr] (#8050) George HotzandGitHub 2024-12-05 11:51:20 +08:00
• 4a208bfb28 bump download cache version geohot 2024-12-05 11:42:34 +08:00
• df18e7cc37 accept filename decorator [pr] (#8049) George HotzandGitHub 2024-12-05 11:40:59 +08:00
• c3187087f7 QwQ-32B-Preview support (#7962) Francis LataandGitHub 2024-12-04 21:46:37 -05:00
• b3220ca7b1 test cases of always True/False lt (#8048) chenyuandGitHub 2024-12-04 20:38:40 -05:00
• 8bb806888b hook_overflow -> safe_exp2 [pr] (#8047) chenyuandGitHub 2024-12-04 19:05:38 -05:00
• 99abdc6d39 minor push_swizzle_down_through_elementwise cleanup [pr] (#8046) chenyuandGitHub 2024-12-04 17:22:37 -05:00
• 5933ec8dc3 use argfix in smax/smin and remove if [pr] (#8045) chenyuandGitHub 2024-12-04 17:06:13 -05:00
• 4e518334b8 minor get_grouped_dims cleanup [pr] (#8044) chenyuandGitHub 2024-12-04 16:22:51 -05:00
• 5ce8090d42 simple onnx_ops cleanups (#8003) 3 people 2024-12-05 04:33:03 +08:00
• 70db1bab5c Fold nested div with const (#8010) 3 people 2024-12-04 20:59:09 +01:00
• 0693158d28 lower v_theoretical gemv on red (#8042) chenyuandGitHub 2024-12-04 13:59:40 -05:00
• 5c2b1089b2 vectorized input in div_and_mod_folding returns None [pr] (#8041) chenyuandGitHub 2024-12-04 13:36:41 -05:00
• ff6def9ffb simple contiguous_while_contiguous prereqs [pr] (#8038) qazalandGitHub 2024-12-04 10:00:28 -05:00
• c9e7701417 Fast YoloV8 on WebGPU (#8036) Ahmed HarmoucheandGitHub 2024-12-04 15:23:09 +01:00
• b116e1511d make device on uop optional [pr] (#8034) qazalandGitHub 2024-12-04 07:18:00 -05:00
• 13eedd373b Run WebGPU tests on ubuntu (#8033) Ahmed HarmoucheandGitHub 2024-12-04 12:42:04 +01:00
• fb89971e73 use BufferedReader (#8032) leopfandGitHub 2024-12-04 12:08:54 +01:00
• 08657cb7b0 hotfix: bump expectations in speed_v_theoretical geohot 2024-12-04 19:00:33 +08:00
• ea65c79ba2 hotfix: don't spam BEAM debug in speed_v_theoretical geohot 2024-12-04 18:47:16 +08:00
• 09b00b1b04 hotfix: use kernel timings instead of python timings in speed_v_theoretical geohot 2024-12-04 18:36:17 +08:00
• 8f65c1fafb simpler block reorder function [pr] (#8031) George HotzandGitHub 2024-12-04 17:57:35 +08:00
• f0401e14e8 tar_extract with Tensors (#7853) 4 people 2024-12-04 10:03:19 +01:00
• 1e06aefde7 bunch up ops for lines [pr] (#8030) George HotzandGitHub 2024-12-04 17:03:01 +08:00
• e9c5b23ba1 Use MTLCompiler directly (v2) (#7920) 3 people 2024-12-04 10:36:48 +02:00
• bb98bae751 local reordering in block (#8029) George HotzandGitHub 2024-12-04 15:11:29 +08:00
• 4cb630ac1c hotfix: early INDEX geohot 2024-12-04 14:47:47 +08:00
• fdd1e56827 clean up rewrite logic + merge siblings (#8026) George HotzandGitHub 2024-12-04 13:26:16 +08:00
• 004b2ecff5 remove lt/gt/le/ge from SimpleMathTrait [pr] (#8027) chenyuandGitHub 2024-12-04 00:24:33 -05:00
• 39e0fc05f5 update function to not use gt/lt [pr] (#8025) chenyuandGitHub 2024-12-03 22:39:06 -05:00
• cfd4d19250 replace .lt in rewrite rules with < [pr] (#8024) chenyuandGitHub 2024-12-03 21:34:47 -05:00
• 0c060fa040 update uop and tests to not use lt/gt/le/ge [pr] (#8023) chenyuandGitHub 2024-12-03 21:02:52 -05:00
• 03bf9c2985 unused mul add lt rule [pr] (#8022) chenyuandGitHub 2024-12-03 19:38:34 -05:00
• 7fda464b08 hcq c-like args state (#8020) nimlgenandGitHub 2024-12-03 23:53:35 +03:00
• 099364ed32 lazy srcs shape mistmatch assert + fix ASSIGN [pr] (#8014) qazalandGitHub 2024-12-03 15:40:37 -05:00
• f14dd1488e reduce on wmma (#8016) ignaciosicaandGitHub 2024-12-03 14:46:28 -03:00
• dacb1ff38a minor nn cleanups (#8018) chenyuandGitHub 2024-12-03 12:34:52 -05:00
• 35c30f76f2 minor tweak in ptx asm_for_op [pr] (#8017) chenyuandGitHub 2024-12-03 12:34:22 -05:00
• a5af4e5596 clean up wgsl_matcher [pr] (#8015) chenyuandGitHub 2024-12-03 11:55:03 -05:00
• db330a3110 Remove WebGL (#8012) Ahmed HarmoucheandGitHub 2024-12-03 16:02:53 +01:00
• ef3752625b add test case of realize_size with 0 in shape (#8011) chenyuandGitHub 2024-12-03 09:19:50 -05:00
• 8818046940 YoloV8 on WebGPU (#8007) Ahmed HarmoucheandGitHub 2024-12-03 15:10:41 +01:00
• 09eac42fd6 cache indexed uops in st [pr] (#8008) George HotzandGitHub 2024-12-03 21:27:07 +08:00
• e44183647f Improved div folding (#7996) 3 people 2024-12-03 14:11:25 +01:00
• 32675a8a77 sacrifice ClangGraph on the altar of lines [pr] (#8009) George HotzandGitHub 2024-12-03 21:11:15 +08:00
• 5441127417 assert const folding return shape matches [pr] (#8006) qazalandGitHub 2024-12-03 06:31:06 -05:00
• dddfb494d7 don't mutate the uop/lazybuffer, just the Buffer [pr] (#8000) George HotzandGitHub 2024-12-03 19:03:51 +08:00
• ba1183314a const_like can return a valid [pr] (#8005) qazalandGitHub 2024-12-03 05:42:12 -05:00
• 4e91533419 test: don't ref until schedule (#8004) qazalandGitHub 2024-12-03 05:06:52 -05:00
• b8bf5b2787 minor uop speedups [pr] (#8002) George HotzandGitHub 2024-12-03 17:04:48 +08:00
• 1028b34a20 add typing to basicblocks (#7999) George HotzandGitHub 2024-12-03 15:05:11 +08:00
• 0905f87b68 hotfix: print only kernel time geohot 2024-12-03 14:25:08 +08:00
• 17d5719a38 add process replay to webgpu tests (#7998) chenyuandGitHub 2024-12-02 20:27:29 -05:00
• c7bc75e634 alu(c?t0:f0, c?t1:f1) -> c?alu(t0,t1):alu(f0,f1) (#7900) chenyuandGitHub 2024-12-02 17:19:27 -05:00
• b91fa24387 script to run regressed sd conv on metal (#7995) chenyuandGitHub 2024-12-02 15:34:27 -05:00
• 0a2e10be1d add SELU to Tensor (#7993) geohotstanandGitHub 2024-12-02 23:04:01 +08:00
• 146e1caea3 Downgrade wgpu to prevent sd segfault (#7969) Ahmed HarmoucheandGitHub 2024-12-02 15:48:44 +01:00
• 077e7e8ed2 fix: private segment sgpr on gfx103x (#7987) 3 people 2024-12-02 07:54:50 -05:00
• bb606e5bcf process replayable ops.py changes from delete_lazy [pr] (#7994) qazalandGitHub 2024-12-02 06:38:31 -05:00
• 0c7477b108 no bool in range [pr] (#7988) George HotzandGitHub 2024-12-02 19:05:16 +08:00
• 8909dbd82c Remove wgpu specific checks from stable diffusion example (#7991) Ahmed HarmoucheandGitHub 2024-12-02 11:31:14 +01:00
• e2916ff210 image dtype fixup refactor for delete_lazy [pr] (#7989) qazalandGitHub 2024-12-02 05:25:13 -05:00
• 5340d3dedf Merge pull request #7986 from tinygrad/atomics-in-smem-wgpu Ahmed HarmoucheandGitHub 2024-12-02 10:38:19 +01:00
• dfae038580 Simplify render_buf_dt Ahmed Harmouche 2024-12-02 10:27:59 +01:00
• 1ea0925744 Support packed types in smem in webgpu Ahmed Harmouche 2024-12-02 09:59:05 +01:00
• 61b2cac507 basicblock is dataclass (#7985) George HotzandGitHub 2024-12-02 16:48:39 +08:00
• 275951b730 clean up a few parents -> toposort [pr] (#7984) George HotzandGitHub 2024-12-02 15:59:31 +08:00
• f17af70d17 replace all sparents with toposort (#7983) George HotzandGitHub 2024-12-02 15:00:30 +08:00
• b09310d8c2 add toposort method to uops, faster linearize [pr] (#7982) George HotzandGitHub 2024-12-02 14:46:16 +08:00
• b797aee720 uop global buf number tracking try 2 [pr] (#7912) qazalandGitHub 2024-12-02 01:45:17 -05:00