mirrors/tinygrad
2
0
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-06-24 02:14:17 +00:00
Code Issues Projects Releases Packages Wiki Activity
13,471 commits 280 branches 15 tags 325 MiB
Commit graph

13,471 commits

This branch
This branch
All branches
Author SHA1 Message Date
George Hotz
263b724143
 		one cache and bump it (#13258) 2025-11-13 07:33:31 -08:00
George Hotz
5efa727b83
 		move _pool to MovementMixins (#13257) 2025-11-13 07:28:52 -08:00
George Hotz
bcdfc109b5 hotfix: disable flaky test 2025-11-13 06:19:28 -08:00
qazal
006dea4c3e
 		roc: only save instruction execs (#13254) 2025-11-13 21:28:40 +08:00
nimlgen
f9586b38ba
 		system: pci mask and val (#13251) 2025-11-13 20:44:58 +08:00
George Hotz
7316da3253
 		new readme (#13250) 
* new readme

* update
 2025-11-13 00:48:28 -08:00
George Hotz
17aa3379e9 hotfix: improve self_tokenize 2025-11-13 00:18:57 -08:00
chenyu
4e5a9132e7
 		JIT_BATCH_SIZE=0 in compile3 (#13245) 
fixed some enqueue time
 2025-11-12 23:12:45 -05:00
wozeparrot
759557f633
 		feat: move tk tests to testextra (#13242) 2025-11-12 17:06:53 -08:00
chenyu
3f939f3d3c
 		update pm_simplify_valid (#13241) 
* update pm_simplify_valid

fixed openpilot conv regression

* IMAGE training is broken
 2025-11-12 19:40:02 -05:00
chenyu
f9851a852f
 		minor update to uop_given_valid [pr] (#13243) 
split from #13241
 2025-11-12 19:03:18 -05:00
qazal
fe2876a6d8
 		hotfix: second GB/s in viz (#13240) 2025-11-13 07:14:27 +08:00
George Hotz
a23dea202b
 		actually make AMD_LLVM not default (#13238) 2025-11-12 15:07:23 -08:00
George Hotz
ab9fa964d8
 		DISABLE_COMPILER_CACHE -> CCACHE (#13234) 
* DISABLE_COMPILER_CACHE -> CCACHE

* Fix cachekey assignment in Compiler constructor
 2025-11-12 15:07:09 -08:00
qazal
be2e24cb25
 		roc: requires sudo to install (#13237) 2025-11-12 16:59:22 -05:00
George Hotz
8f1f195b6d hotfix: no hexdump for usbgpu patch.py 2025-11-12 12:05:37 -08:00
nimlgen
9a53fcbde4
 		amd: sqtt on rdna3.5 (#13233) 2025-11-13 03:30:42 +08:00
George Hotz
13f10a31dc
 		AMD_LLVM default off (#13232) 2025-11-12 11:06:33 -08:00
qazal
8b26cf2b3d
 		sqtt: update rcp timing test (#13231) 
* sqtt: assert correct output in timing test

* found why
 2025-11-13 02:01:54 +08:00
Jan Akhremchik
bc8e537423
 		Add NONZERO op to onnx backend (#13211) 2025-11-12 08:55:51 -08:00
nimlgen
af17e07251
 		viz: sqtt touchups (#13228) 
* viz: sqtt touchups

* revert

* matches
 2025-11-12 22:40:37 +08:00
qazal
7a6853fa40
 		viz: show python callstack in the first graph (#13218) 2025-11-12 20:52:28 +08:00
nimlgen
82eb63d3ad
 		qcom: auto switch idle timer when profiling (#13230) 
* qcom: auto switch idle timer when profiling

* fi
 2025-11-12 20:31:24 +08:00
nimlgen
fcd8d0751a
 		test_timing for hip (#13229) 2025-11-12 20:28:58 +08:00
qazal
74b9d33acb
 		viz: direct link to program source (#13227) 2025-11-12 16:27:13 +08:00
wozeparrot
371c1f2355
 		tk: move tiles to class (#13224) 2025-11-11 21:53:46 -08:00
Christopher Milan
41a098a82d
 		In-tree autogen: libc.py (#13217) 
* checkout changes from autogen branch

* parents

* pylint happy

* move sys to system in helpers.py

* typo

* typo
 2025-11-11 19:13:48 -08:00
wozeparrot
222bb12ddf
 		tk softmax (#13205) 2025-11-11 15:13:16 -08:00
wozeparrot
787f0070ed
 		feat: don't use output reg as local reduce reg (#13203) 2025-11-11 14:35:16 -08:00
chenyu
ece1415def
 		clean up image_dot and image_conv2d (#13222) 
* clean up image_dot and image_conv2d

* those are fine

* interesting
 2025-11-11 15:53:03 -05:00
nimlgen
2f0ea29b34
 		qcom: 48bit timestamps (#13214) 
* qcom: 48bit timestamps

* f

* lol

* fix
 2025-11-12 04:14:33 +08:00
qazal
bc55bc4849
 		cleanup test_viz profiler tests (#13221) 2025-11-12 03:46:48 +08:00
chenyu
23b90945c3
 		add a benchmark for openpilot vision with DEBUG=2 (#13219) 
see per kernel speed, also disable the jobs for 0.9.9
 2025-11-11 14:41:52 -05:00
George Hotz
c2075f3613
 		gc disable during big rewrites (#13215) 
* gc disable during big rewrites

* cleaner with helper
 2025-11-11 10:30:47 -08:00
Roelof van Dijk
e59313da08
 		migrate pytest and ruff (#13216) 2025-11-11 13:27:51 -05:00
Gaétan Lepage
6fd7ce3832
 		migrate to pyproject.toml (#13189) 
* migrate to pyproject.toml

* move mypy config to pyproject.toml
 2025-11-11 09:09:27 -08:00
qazal
8002921a04
 		viz: improve the program run tooltip (#13212) 
* add tflops to tooltip format

* show if the run was batched
 2025-11-12 00:56:03 +08:00
qazal
f91e366a17
 		viz: display the graph layout recursion error (#13194) 
* viz: display the graph layout recursion error

* share styles

* +min-width

* same thing

* inline the append
 2025-11-11 15:25:12 +08:00
wozeparrot
73497af4c0
 		clean: use np for allclose (#13204) 2025-11-10 23:02:43 -08:00
George Hotz
a6360fd94d
 		store can have shape (#13202) 
* store can have shape

* _shape
 2025-11-10 22:16:47 -08:00
b1tg
f3692b7406
 		clean up hip renderer (#13063) 
* clean up hip renderer

* ocml

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2025-11-11 00:44:24 -05:00
chenyu
22b8579234
 		one last regressed dm kernel (#13201) 2025-11-10 23:30:52 -05:00
chenyu
58b7e4fab3
 		GROUPTOP heuristic on more axes (#13206) 
fixed dm speed
 2025-11-10 23:30:37 -05:00
chenyu
829cdafccc
 		update openpilot slow conv uop ast (#13197) 
the two remaining slow ones
 2025-11-10 17:03:20 -05:00
George Hotz
0c978d45e6
 		stub attention (#13196) 
* stub attention

* name the kernels
 2025-11-10 13:48:38 -08:00
chenyu
58c30fc7ce
 		minor image_conv2d cleanup (#13193) 2025-11-10 16:05:40 -05:00
chenyu
60e55d9a2d
 		line count 18500 (#13191) 2025-11-10 13:52:13 -05:00
nimlgen
09a59c2203
 		qcom: support new chip versioning (#13185) 
* qcom: support new chip versioning

* ops

* nit

* fix

* f
 2025-11-10 23:57:29 +08:00
qazal
50934050bc
 		sqtt: append all wave execs (#13190) 2025-11-10 23:50:08 +08:00
qazal
38a24731a1
 		cleanup sqtt tooling (#13188) 
* cleanup viz/serve.py

* use latest profile in rgptool.py

* unwrap nullable in roc.py, fix disasms typing
 2025-11-10 20:52:57 +08:00