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

• a352a6938f simplify group_for_reduces in get_index [pr] (#7851) chenyuandGitHub 2024-11-22 11:53:21 -05:00
• af5d77f684 move sint_to_uop from view.py to ops.py [pr] (#7848) chenyuandGitHub 2024-11-22 11:15:02 -05:00
• f6d1201c48 variable_to_uop -> sint_to_uop [pr] (#7847) chenyuandGitHub 2024-11-22 10:54:59 -05:00
• 40d7535eeb clean up DTYPES_DICT [pr] (#7845) chenyuandGitHub 2024-11-22 10:01:34 -05:00
• 4453ab51e1 use ceildiv in View.stride [pr] (#7844) chenyuandGitHub 2024-11-22 08:38:05 -05:00
• 9828277c03 view doesn't have buffer, fix the tests [pr] (#7841) qazalandGitHub 2024-11-22 07:41:55 -05:00
• 7e8777eee9 faster assign scheduling [pr] (#7839) qazalandGitHub 2024-11-22 06:23:59 -05:00
• 6229d87f45 simpler reshape symbolic shape check [pr] (#7837) chenyuandGitHub 2024-11-21 22:53:57 -05:00
• 1d6d842887 move DSP to extra (room for webgpu) [pr] (#7836) George HotzandGitHub 2024-11-22 11:32:57 +08:00
• 8ff6cba9f0 simpler swizzle_r new_axis [pr] (#7835) chenyuandGitHub 2024-11-21 22:26:41 -05:00
• 6fc7013463 put all DSP in dsp file [pr] (#7833) George HotzandGitHub 2024-11-22 11:22:59 +08:00
• e39af63156 no loop assert in ops_python [pr] (#7834) George HotzandGitHub 2024-11-22 11:17:36 +08:00
• d18b948f48 ptxcompiler isn't a cudacompiler [pr] (#7832) George HotzandGitHub 2024-11-22 10:57:22 +08:00
• 855f9a767a add restype for msg method for type annotation and consistency (#7828) 3 people 2024-11-22 09:17:58 +08:00
• d5c9fafff5 default run stable diffusion benchmark with fp16 (#7831) chenyuandGitHub 2024-11-21 15:58:17 -05:00
• 69e382216d fix wino conv output dtype for half inputs (#7829) chenyuandGitHub 2024-11-21 12:13:54 -05:00
• cf1ec90ad4 add inverse trig functions to Tensor (#7805) 3 people 2024-11-21 22:13:36 +08:00
• 5399ff6d06 add UOp.const_with_shape [pr] (#7825) qazalandGitHub 2024-11-21 08:13:23 -05:00
• 2f884b2384 good suggestions from mypy lineprecision-report for schedule.py [pr] (#7823) qazalandGitHub 2024-11-21 06:59:51 -05:00
• e378aeb94e assert view degrade to const tests post scheduler graph_rewrite [pr] (#7822) qazalandGitHub 2024-11-21 06:00:41 -05:00
• cdc431803f early mark uops as realized [pr] (#7821) qazalandGitHub 2024-11-21 05:02:59 -05:00
• 4542c0f000 get buffer size just from Ops.BUFFER [pr] (#7820) qazalandGitHub 2024-11-21 04:00:18 -05:00
• 877b440fde derive device (dname) from UOp [pr] (#7819) qazalandGitHub 2024-11-21 03:38:22 -05:00
• 75c082b883 move CONST/BIND -> VALID to matchers (#7818) qazalandGitHub 2024-11-21 03:07:01 -05:00
• df6f1815ad remove jit_cache from self in GraphRunner [pr] (#7817) George HotzandGitHub 2024-11-21 13:26:37 +08:00
• e9ae2ccd09 _prg to match _buf [pr] (#7816) George HotzandGitHub 2024-11-21 12:44:48 +08:00
• 439911b2e6 disable disable_abstract_method [pr] (#7815) George HotzandGitHub 2024-11-21 12:28:57 +08:00
• c5d458ce02 BufferSpec and ProgramSpec [pr] (#7814) George HotzandGitHub 2024-11-21 12:18:05 +08:00
• 490a6130af more hcq typing [pr] (#7813) George HotzandGitHub 2024-11-21 11:23:07 +08:00
• 9df5a62c5e unify to HWQueue [pr] (#7812) George HotzandGitHub 2024-11-21 10:33:08 +08:00
• 11cea00090 lower vs_theoretical conv tflops threshold for nv (#7811) chenyuandGitHub 2024-11-20 20:03:49 -05:00
• 46aa23539f generate and print mypy lineprecision report (#7809) chenyuandGitHub 2024-11-20 16:53:17 -05:00
• c815d7b56e run bfloat16 tensor core in metal benchmark (#7808) chenyuandGitHub 2024-11-20 15:34:07 -05:00
• 33a496279b load_state_dict check v.shape instead of v.lazydata.shape (#7807) chenyuandGitHub 2024-11-20 14:39:30 -05:00
• fc3154a7b3 metal bf16 tc support [pr] (#7408) ignaciosicaandGitHub 2024-11-20 16:39:08 -03:00
• 66a069ee25 add replicate mode to Tensor.pad (#7802) geohotstanandGitHub 2024-11-20 21:39:58 +08:00
• eb0bb7dc0b final dname to device [pr] (#7806) George HotzandGitHub 2024-11-20 20:20:28 +08:00
• bc977fec53 dname -> device [pr] (#7804) George HotzandGitHub 2024-11-20 17:57:14 +08:00
• 0a74acd90e add proper typing to HCQ [pr] (#7803) George HotzandGitHub 2024-11-20 17:20:39 +08:00
• 6688539bc9 rename device to dev so Buffer can be Allocator [pr] (#7799) George HotzandGitHub 2024-11-20 15:47:26 +08:00
• 9adeb1041c fix advanced setitem with 1 in shape (#7797) ttomsaandGitHub 2024-11-20 01:04:59 +00:00
• d800a79112 use "signed char" for int8 (#7796) chenyuandGitHub 2024-11-19 19:29:54 -05:00
• f16122f9c4 update README to make it runs with just tinygrad (#7795) chenyuandGitHub 2024-11-19 17:25:12 -05:00
• 170ece6605 fix advanced setitem overlap with 0 (#7793) ttomsaandGitHub 2024-11-19 21:03:55 +00:00
• 159c0bf25e test_kernel_cache_in_action: fix test (#7792) Gaétan LepageandGitHub 2024-11-19 19:34:56 +01:00
• 913a27ee27 from_buffer on metal was never called [pr] (#7791) George HotzandGitHub 2024-11-20 00:35:17 +08:00
• 56017c52a0 Raise error when model architecture does not match state dict (#7772) Eitan TurokandGitHub 2024-11-19 11:11:54 -05:00
• d71fe7faa5 rename allocator methods to not conflict [pr] (#7788) George HotzandGitHub 2024-11-20 00:10:29 +08:00
• d5f76462c8 fix CI beautiful_mnist dir (#7790) chenyuandGitHub 2024-11-19 09:59:02 -05:00
• aeaf574a05 add failure test for setitem bug (#7786) geohotstanandGitHub 2024-11-19 21:54:21 +08:00
• 1e31b5ba6b hotfix: ctx doesn't impact process replay [pr] (#7785) qazalandGitHub 2024-11-19 14:17:01 +02:00
• 8360bbd88d faster assign view check [pr] (#7781) qazalandGitHub 2024-11-19 13:42:51 +02:00
• 3daa376107 remove numpy from assign [pr] (#7784) George HotzandGitHub 2024-11-19 19:34:53 +08:00
• fbb4099b3c add test for compile3 [pr] (#7783) 3 people 2024-11-19 19:26:51 +08:00
• 4f6071d919 capture the schedule context in process replay [pr] (#7782) qazalandGitHub 2024-11-19 13:12:00 +02:00
• f493d480e3 metadata appending to graph_rewrite (#7780) qazalandGitHub 2024-11-19 12:05:42 +02:00
• 73ea913050 really not using numpy in gpt2 example (#7779) chenyuandGitHub 2024-11-18 23:21:16 -05:00
• e6debda5c4 remove numpy from gpt2 and llama examples (#7778) chenyuandGitHub 2024-11-18 22:48:17 -05:00
• 005636304b have VIZ=1 use HTTP/1.1 for keep-alive [pr] (#7776) George HotzandGitHub 2024-11-19 09:38:12 +08:00
• 65f188aafb bump version to 0.10.0 v0.10.0 geohot 2024-11-19 08:27:28 +08:00
• 26200574dc load_state_dict test cases when model and data shard differently (#7774) chenyuandGitHub 2024-11-18 16:08:24 -05:00
• a1c1b9547f Context manager support for tqdm (#7770) Francis LataandGitHub 2024-11-18 14:12:03 -05:00
• 8100109c9d Add replicate mode to Tensor.pad (#7608) 3 people 2024-11-18 23:55:38 +08:00
• 62db6398a5 delete buffer tracking from ScheduleContext [pr] (#7766) qazalandGitHub 2024-11-18 16:47:32 +02:00
• ed76d3ceac Fix AMD queue CWSR memory size (#7765) ShuniandGitHub 2024-11-18 15:22:03 +01:00
• f02462c5cb swizzle tc [pr] (#7633) 3 people 2024-11-18 10:08:21 -03:00
• 6ea4a173e7 make is_realized a property [pr] (#7763) qazalandGitHub 2024-11-18 13:15:37 +02:00
• 5de0ea40f3 reorder Tensor.__init__ to match type (#7758) chenyuandGitHub 2024-11-17 21:32:48 -05:00
• 66d7d5af50 fix Tensor(MultiLazyBuffer) with different dtype should fail (#7757) chenyuandGitHub 2024-11-17 21:05:45 -05:00
• b1d734a02c remove the -1 then -(-1) in Tensor.argmax (#7753) chenyuandGitHub 2024-11-17 16:54:09 -05:00
• e3081355fe minor Tensor.einsum cleanup (#7752) chenyuandGitHub 2024-11-17 16:11:30 -05:00
• 8b08a72657 consmetic change to Tensor._pool (#7751) chenyuandGitHub 2024-11-17 15:38:11 -05:00
• df817297b6 fix passing acc_dtype="" to Tensor.prod should fail (#7750) chenyuandGitHub 2024-11-17 11:38:13 -05:00
• 55707fd00d fix passing sum_acc_dtype="" to Tensor.sum should fail (#7748) chenyuandGitHub 2024-11-17 10:58:41 -05:00
• f18296e23c simpler Tensor._reduce (#7747) chenyuandGitHub 2024-11-17 09:20:00 -05:00
• 0cc8de2f15 reverse map buf_uops [pr] (#7743) qazalandGitHub 2024-11-17 15:29:56 +02:00
• 0292ae7508 Tensor.meshgrid cleanup (#7741) chenyuandGitHub 2024-11-17 08:26:53 -05:00
• 40642cb9ea to_uop split paths part 2 [pr] (#7746) qazalandGitHub 2024-11-17 15:07:28 +02:00
• 99024b922b to_uop one path for all ops part 1 (#7745) qazalandGitHub 2024-11-17 14:12:44 +02:00
• eeb222f98b add UOp.new_buffer [pr] (#7742) qazalandGitHub 2024-11-17 10:44:52 +02:00
• a15a900415 fix Tensor.meshgrid for 1D input and check indexing (#7740) chenyuandGitHub 2024-11-16 23:39:30 -05:00
• 72a41095bc add Tensor.meshgrid (#7714) 3 people 2024-11-17 12:06:47 +08:00
• 1a5e896bd4 [pr] Have PTX share code with LLVM (#7635) 3 people 2024-11-17 10:53:56 +08:00
• f2f7384b67 _resolve_dim cleanup (#7736) chenyuandGitHub 2024-11-16 11:05:39 -05:00
• e777211a00 Tensor.repeat cleanup (#7735) chenyuandGitHub 2024-11-16 10:43:45 -05:00
• f1efd84c92 fix repeat_interleave with negative dim (#7734) chenyuandGitHub 2024-11-16 10:15:29 -05:00
• e3105675fb cond.where(True, False) is cond (#7733) chenyuandGitHub 2024-11-16 09:44:17 -05:00
• 40ae0e9115 smaller big graph (#7695) qazalandGitHub 2024-11-16 16:04:57 +02:00
• f3f95ab9d9 flatten fusion upats [pr] (#7732) qazalandGitHub 2024-11-16 15:26:19 +02:00
• ec8c5598f6 refactor to generic UPat for sourcing unrealized bufs [pr] (#7731) qazalandGitHub 2024-11-16 15:01:22 +02:00
• 597a239e28 Remove UnaryOps, BinaryOps, TernaryOps, MetaOps [pr] (#7725) 3 people 2024-11-16 09:56:56 -03:00
• 22da31b223 clean up Tensor.dot (#7728) chenyuandGitHub 2024-11-15 18:21:15 -05:00
• 4338c450ac fix max_pool2d for int tensor with padding (#7726) chenyuandGitHub 2024-11-15 16:22:11 -05:00
• d736ae7153 example script to show BasicTransformerBlock speed regression (#7724) chenyuandGitHub 2024-11-15 15:48:25 -05:00
• aeb1301bab enable a few tests that work now (#7721) chenyuandGitHub 2024-11-15 14:30:52 -05:00
• fc1e123138 minor cleanup in lazy.py (#7719) ignaciosicaandGitHub 2024-11-15 15:48:24 -03:00
• ef4f402946 add property to flag contig buffer uop [pr] (#7716) qazalandGitHub 2024-11-15 16:27:47 +02:00
• 313af6d23c assert buffer VIEW is void [pr] (#7715) qazalandGitHub 2024-11-15 16:02:59 +02:00
• c37d142cf8 Refactor metal tc wmma kernel rendering (#7416) ignaciosicaandGitHub 2024-11-15 10:23:08 -03:00
• bddee26114 Ops.VALID cleanup, move recursive tests [pr] (#7713) qazalandGitHub 2024-11-15 14:22:46 +02:00