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

• 03994e0011 load torch files without torch geohot 2020-11-21 13:43:53 -08:00
• 26899869a2 Update tensor.py (#128) Marcel BischoffandGitHub 2020-11-21 12:16:03 -05:00
• f190ca446d Detach (#123) 3 people 2020-11-20 03:03:42 +00:00
• 8383ff40ad fix pyopencl (#125) Colin MankoandGitHub 2020-11-19 22:03:04 -05:00
• 5797e63d9b Train efficientnet should respect NUM environment variable (#122) 3 people 2020-11-17 04:02:31 +00:00
• ee99d016e9 tensor implementation for rmsprop and adam (#121) dustcollector12andGitHub 2020-11-17 00:07:49 +01:00
• 17bf90dbe4 unbroadcasting works on the GPU geohot 2020-11-16 09:16:55 -08:00
• 17eab716b6 unbroadcast GPU template geohot 2020-11-16 08:16:36 -08:00
• 2ffb8de1ea move efficientnet to extra geohot 2020-11-16 08:08:07 -08:00
• 13d34373d1 move gradcheck to extra, clean up unbroadcast geohot 2020-11-16 08:03:31 -08:00
• ed4c35e2e9 channels on the inside geohot 2020-11-15 21:19:59 -08:00
• fb1df81c7d Fix train_efficientnet (#120) 3 people 2020-11-16 04:50:31 +00:00
• 1207fe4c7d cleanup LogSoftmax geohot 2020-11-15 20:49:57 -08:00
• d1441de3a6 minor cleanups geohot 2020-11-15 20:39:19 -08:00
• 37a210f868 touchups and lines geohot 2020-11-15 20:26:52 -08:00
• 5ea3d76dfb Topological sort, zero_grads (#119) 3 people 2020-11-16 04:25:29 +00:00
• a35425189d binop fast path for no broadcast geohot 2020-11-15 19:12:14 -08:00
• c7b7f8ccc8 Backwards ops supporting broadcasting (#118) 3 people 2020-11-15 18:21:10 -05:00
• 55d93017e4 Simplify more (#117) 3 people 2020-11-14 14:15:31 +00:00
• 28474949b8 refactoring of forward in reshape (#115) dustcollector12andGitHub 2020-11-13 22:20:43 +01:00
• 6f033ea30a enable local images for efficientnet.py (#116) dustcollector12andGitHub 2020-11-13 16:00:12 +01:00
• 420af82888 General broadcasting of binary operations (#114) 3 people 2020-11-12 22:27:48 -08:00
• 2b1286eef6 Don't wrap np.int32 in a function, use an alias (#113) damianzimandGitHub 2020-11-13 04:32:19 +01:00
• 08aa60d9d0 broadcasting 1s at the start, 1 kernel/4 divs version (#110) 3 people 2020-11-12 21:33:35 +00:00
• f773ef3996 ⚡ tanh non first class op (#111) 3 people 2020-11-13 03:02:50 +05:30
• 608bdd4872 adds broadcasting test cases (#106) Ryan NephandGitHub 2020-11-12 07:08:28 -08:00
• f1d21afe88 Somewhat more generic broadcasting (#105) 3 people 2020-11-12 04:33:00 +00:00
• 8827a536e0 GPU MaxPool2D.backward(); TinyConvNet train passes (#103) Ryan NephandGitHub 2020-11-11 07:58:43 -08:00
• a3989f9e18 Supporting .png files in efficientnet (#102) Marcel BischoffandGitHub 2020-11-10 23:06:24 -05:00
• d93cd945aa reshape makes copies geohot 2020-11-10 16:18:59 -08:00
• d1284fa817 stride tests and i32 geohot 2020-11-10 16:10:14 -08:00
• 7bb803c5e0 Conv2D backward on GPU (#93) Marcel BischoffandGitHub 2020-11-10 19:07:33 -05:00
• 5577b9d3a0 clean up imports geohot 2020-11-10 15:53:05 -08:00
• db755fa103 promote swish to a tensor ops geohot 2020-11-10 15:48:11 -08:00
• 5f4b76a21b touch ups geohot 2020-11-10 15:44:47 -08:00
• 52ee913c98 move the mnist loader out of tinygrad proper geohot 2020-11-10 15:37:39 -08:00
• 498b4d2f27 i32 and reduce line count a bit geohot 2020-11-10 15:35:30 -08:00
• df64658a2c weee, opencl tests in CI geohot 2020-11-10 10:04:45 -08:00
• d47a128812 pocl geohot 2020-11-10 10:02:13 -08:00
• c05401a9ca sudo maybe geohot 2020-11-10 09:53:49 -08:00
• 09bc8eddfe clinfo geohot 2020-11-10 09:51:38 -08:00
• 58e703d099 fix tests geohot 2020-11-10 09:49:19 -08:00
• 23405cec43 intel opencl geohot 2020-11-10 09:41:40 -08:00
• 33090c4b0d install more geohot 2020-11-10 09:34:56 -08:00
• a52590e76c cpu opencl maybe geohot 2020-11-10 09:32:54 -08:00
• f513302955 refactor profiler geohot 2020-11-10 07:31:16 -08:00
• f27628b21c No separate pad2d kernel needed (#99) 3 people 2020-11-10 14:47:53 +00:00
• 2d4a5d5950 readme geohot 2020-11-10 01:27:04 -08:00
• 6e6bcbe5f2 shapes on backward geohot 2020-11-10 01:23:22 -08:00
• 56f71ae8e5 Cleanup (#96) Ryan NephandGitHub 2020-11-10 01:20:04 -08:00
• 55012d21bb debug in backward pass too geohot 2020-11-10 01:19:52 -08:00
• 5d1985312c miniprofiler is real geohot 2020-11-10 01:05:29 -08:00
• c76a20b4be 4s and 7s work geohot 2020-11-10 00:54:17 -08:00
• ae0cd17c2d debug is env var, and simpler faster pad2d geohot 2020-11-10 00:42:23 -08:00
• f7d10d5639 DEBUG flag geohot 2020-11-10 00:36:59 -08:00
• 6a56d5d030 remove pyopencl geohot 2020-11-09 23:58:35 -08:00
• 943ff6490d logic error okay too geohot 2020-11-09 23:55:21 -08:00
• 29b22d117f will gpu tests work? geohot 2020-11-09 23:51:29 -08:00
• abbf0d1328 cleanup logsoftmax geohot 2020-11-09 23:47:04 -08:00
• aeb90226a8 reduce lines with reduce_op geohot 2020-11-09 23:36:44 -08:00
• 55c914912d minor cleanup geohot 2020-11-09 23:14:59 -08:00
• d41ad2bf37 uint2 saves lines geohot 2020-11-09 23:11:20 -08:00
• 8b23033fa9 support all the enet sizes geohot 2020-11-09 18:04:16 -08:00
• 9db95ab942 fix enet padding geohot 2020-11-09 17:56:57 -08:00
• 866b759d3b match torch api for pad2d geohot 2020-11-09 17:48:56 -08:00
• daf073535f new -> zeros geohot 2020-11-09 17:31:42 -08:00
• 16d564a53c finish unsupporting strided pool, add global avg pool test (#92) Ryan NephandGitHub 2020-11-09 17:31:22 -08:00
• 7ac1b163a5 add backward to enet train geohot 2020-11-09 16:05:52 -08:00
• 8ca9c0205f train_efficientnet is broken still geohot 2020-11-09 16:01:16 -08:00
• 870b84a893 test pad2d backward on GPU geohot 2020-11-09 15:50:37 -08:00
• b541c05d88 Pad2d backward pass on GPU (#89) 3 people 2020-11-09 23:49:37 +00:00
• e46d122f65 not supporting stride geohot 2020-11-09 15:06:58 -08:00
• c21c2a0b62 revert b0c0c5d: Strided Pool funcs (#74) (#87) 3 people 2020-11-09 14:58:18 -08:00
• 53157fb876 add back scale geohot 2020-11-09 10:20:56 -08:00
• 3ffbd47335 Revert "Revert "pygame is fine, cv2 can also do the trick (#79)" (#85)" geohot 2020-11-09 10:18:48 -08:00
• 6b982621f8 Revert "pygame is fine, cv2 can also do the trick (#79)" (#85) George HotzandGitHub 2020-11-09 10:03:38 -08:00
• e7f2f43331 pygame is fine, cv2 can also do the trick (#79) dustcollector12andGitHub 2020-11-09 19:02:06 +01:00
• 7e515308a5 label op subtests by params (#83) Ryan NephandGitHub 2020-11-09 06:25:06 -08:00
• 5bedf566d1 tests should use rtol unless special case (#82) Ryan NephandGitHub 2020-11-08 17:25:11 -08:00
• 04b9312a34 Fix GPU Pooling bug at boundary + better Pooling test coverage (#81) Ryan NephandGitHub 2020-11-08 17:25:01 -08:00
• c57b1b9e7d deleted unnecessary import in utils (#78) niclaswueandGitHub 2020-11-09 00:55:16 +01:00
• b0c0c5d0d6 strided Pool funcs (#74) Ryan NephandGitHub 2020-11-08 11:45:55 -08:00
• 06504a5824 bump version geohot 2020-11-08 09:34:07 -08:00
• db3eccc16b implemented backward for Pad2D & test (#73) ziofilandGitHub 2020-11-08 06:58:42 +01:00
• 75d69e956f readme more geohot 2020-11-07 21:58:20 -08:00
• 35a5c82a2a done as he said (#71) Dimitar VagalinskiandGitHub 2020-11-08 04:28:39 +02:00
• 5265f6c578 add AvgPool2D backward pass on GPU (#68) Ryan NephandGitHub 2020-11-07 12:27:29 -08:00
• ce6c408d78 readmee geohot 2020-11-07 12:26:57 -08:00
• 3ab8fb2189 oops, timing geohot 2020-11-07 12:15:52 -08:00
• e4c5fb0219 webcam support geohot 2020-11-07 12:15:16 -08:00
• 5486135f2d readme geohot 2020-11-07 11:41:27 -08:00
• 9b8fefe4e3 fix broadcast bugggg geohot 2020-11-07 11:38:56 -08:00
• b1ca4dd327 who loves speeeeed geohot 2020-11-07 11:36:52 -08:00
• e6c8321e5b enet is running on GPU geohot 2020-11-07 11:33:46 -08:00
• 30442a086a some broadcasting, pool test is fail geohot 2020-11-07 11:29:42 -08:00
• 7c630ca355 GPU {Avg,Max}Pool2D forward pass (#66) Ryan NephandGitHub 2020-11-07 11:13:24 -08:00
• 40c538c299 fix reshape geohot 2020-11-07 11:10:29 -08:00
• 19c2950310 refactor ops to prepare for broadcasting geohot 2020-11-07 11:02:45 -08:00
• 94d44c97bf add pad2d on GPU geohot 2020-11-07 10:46:36 -08:00
• fbff6ab2e5 fix strided convs, GPU env var for enet geohot 2020-11-07 10:26:37 -08:00