github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-01-14 17:38:06 -05:00
Code Issues Packages Projects Releases Wiki Activity
11,106 Commits 203 Branches 14 Tags
python_speed
Commit Graph

4667 Commits

Author SHA1 Message Date
qazal
3bd992dc95 multi stage graph_rewrite_map (#9803) 
* multistage graph_rewrite_map

* s/merge_map/input_map

* build up kernel_map from the tensor_map
 2025-04-09 15:59:45 +08:00
George Hotz
78caf55154 Revert "FP8 support on NVIDIA (#8631)" 
This reverts commit 2c8e4ea865.
 2025-04-09 12:27:41 +08:00
George Hotz
d1505137ad Revert "move TestOpsFp8s skipTest (#9797)" 
This reverts commit a3aaf92b21.
 2025-04-09 12:27:40 +08:00
chenyu
a3aaf92b21 move TestOpsFp8s skipTest (#9797) 
so get_available_devices is not called when running other tests
 2025-04-08 22:44:07 -04:00
pkotzbach
2c8e4ea865 FP8 support on NVIDIA (#8631) 
* squashed fp8 commits

* tensorcore start

* minor changes

* pre-commit

* pylint

* Delete fp8mul.cu

* clean

* small bugfix

* fix test_dtype

* fix test_dtype_alu

* add EMULATE_CUDA_SM89

* fix ci

* fix test_linearizer

* fix test_linearizer

* fix swizzle

* add debug to simple_matmul

* fixed swizzle

* python emulator

* refactor python emulator

* setup fix

* numpy setup

* ml_dtypes only in emulate_cuda_sm89

* fix pylint

* fix tests

* fix mypy

* fix mypy

* fix ruff

* done python emulator

* add acc type

* tests

* mypy

* clean code

* add cuda tensor core tests to CI

* minor fix

* clean test_dtype.py

* clean cstyle.py

* clean test_ops.py

* fix test

* fix test

* whitespaces

* pylint

* pylint

* amd?

* amd?

* amd

* reduce lines

* mockgpu remove

* fix

* ruff

* ruff

* fix mypy

* ruff

* test only for cuda

* fixed formatting

* small fixes

* small fix

* least_upper_dtype if fp8s not supported

* log and reciprocal are supported for fp8s

* ops python fixes

* dtypes.fp8s use

* e4m3 + e5m2 result dtype test

* truncate linter fix

---------

Co-authored-by: pkotzbach <pawkotz@gmail.com>
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
Co-authored-by: chenyu <chenyu@fastmail.com>
 2025-04-08 21:54:04 -04:00
qazal
f13e9cf2d9 move view_left to grouper.py + tiny reorders [pr] (#9780) 
* move view_left to grouper.py [pr]

* reorder grouper

* test_schedule
 2025-04-08 15:39:28 +08:00
chenyu
7a28133b37 failed test for single softmax backward (#9778) 
getting RecursionError with DONT_GROUP_REDUCES=1
 2025-04-08 02:36:32 -04:00
George Hotz
fefee5d3ab single kernel softmax (#9776) 
* real single kernel softmax

* cleanup

* fix blockend insertion

* add to bert test
 2025-04-08 12:35:48 +08:00
qazal
9963bb51e0 grouper tests cleanups [pr] (#9777) 
* grouper tests cleanups [pr]

* viz

* tuple

* whitespace
 2025-04-08 12:33:11 +08:00
George Hotz
db22094d35 hotfix: update softmax fusion test 2025-04-08 11:23:19 +08:00
Eitan Turok
bb7922b95f Vectorize Transcendental Regression Tests (#9753) 
* init test

* cleanup
 2025-04-08 01:27:39 +08:00
Sieds Lykles
07d1aefaf4 fast idiv (#9755) 
* fast idiv with tests and fuzzer

* Add todo comment

* Add env variable to toggle fast_idiv

* Move env check

* Add fuzz fast_idiv to ci

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2025-04-07 08:32:24 -04:00
nimlgen
fa888ee077 minor test cleanups (#9770) 
* fix test_graph on max

* pcie5
 2025-04-07 15:29:12 +03:00
qazal
891322fd51 split into grouper.py (#9768) 
* split into grouper.py

* update tests

* reorder
 2025-04-07 18:40:59 +08:00
qazal
8ddb1357c0 fix UPat.location after pickle (#9763) 
* fix UPat.location after pickle [pr]

* named upat test
 2025-04-07 15:16:42 +08:00
chenyu
b190d85ad7 benchmark script bert softmax (#9759) 2025-04-07 00:31:18 -04:00
Ignacio Sica
58785181a8 AMD bf16xf32 TC (#9717) 
* dont test bf16 for emulated amd tc

* skip bf16 tc test in ci

* skip bf16 for AMD in test_tensor_cores_codegen

* add simple bf16 gemm test to benchmark
 2025-04-07 11:41:04 +08:00
chenyu
43e4565148 weighted linear in external_benchmark_bert_matmuls (#9757) 
include the linear to get qkv, and permute so that stride matches with the real run
 2025-04-06 23:35:42 -04:00
chenyu
8a585dc5c1 benchmark script for matmuls in bert (#9752) 
2 main matmuls in the bert layers. getting these to be fast makes bert fast
 2025-04-06 19:34:25 +08:00
nimlgen
5f7c79676f jit: prune independent copies (#9749) 
* jit: prune independent copies

* linter

* check kernel cnt
 2025-04-05 20:50:28 +03:00
nimlgen
c2573b247c jit: rename optimize_weights -> replan_buffers_memory_layout (#9751) 2025-04-05 20:35:15 +03:00
chenyu
407ca54382 symbolic fold double where (#9436) 
* symbolic fold double where

a.where(b.where(c, d), d) -> (a & b).where(c, d). a pattern in optimizer

* test case
 2025-04-05 05:12:17 -04:00
Sieds Lykles
9c2fc695b5 cond.logical_not().where(a,b) -> cond.where(b,a) (#9741) 
* Add rule for negation in where, simplifies arange patterns

* 0 becomes 0.0 again

* Only if cond is bool

* ne is never None

* Add a test

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2025-04-04 19:13:32 -04:00
chenyu
fe998798fb linearizer failure test for OUT OF BOUNDS ACCESS (#9738) 2025-04-04 03:48:43 -04:00
George Hotz
8b5a523743 fix minimum length in pattern matcher (#9736) 2025-04-04 14:57:01 +08:00
George Hotz
926b0bcc57 cache folded upcast [pr] (#9733) 2025-04-04 11:23:19 +08:00
George Hotz
cac8bcf8b5 use Ops.REDUCE (#9721) 
* decrease bert python time [pr]

* order copies

* Revert "order copies"

This reverts commit 3f62c8693b.

* rewrite count

* Ops.REDUCE

* acc first in the add chain

* Fix tensor core acc

* arange patterns look good

* fix multireduce gate

* reduce rewrite rule

* bump that to 15 minutes

* multiwmma isn't fusing

* gep through wmma is gep pushing

* bump that timeout too, it's all env setup

* add failing test
 2025-04-04 10:14:34 +08:00
nimlgen
949459fdd6 jit: fix deallocate on unallocated buffers in free_intermediates (#9699) 2025-04-03 18:32:51 +03:00
geohotstan
ac713e04db ONNX add output shape validation (#9720) 
* add output shape validation and remove support for sequence_type

* nit better err msg

* add sequence_type back

* improve err msg

* Revert "improve err msg"

This reverts commit dc9eaea4bb.

* Revert "add sequence_type back"

This reverts commit 288170b2d9.

* do explicit shape equality

* small nit
 2025-04-03 05:44:53 -04:00
chenyu
79145e3d40 cleanup truncate_bf16 [pr] (#9725) 
use torch bfloat16 for groundtruth in test. also a TODO for discrepancy
 2025-04-03 05:43:49 -04:00
Ignacio Sica
bc2d86195e increase test tolerance (#9719) 2025-04-03 15:24:09 +08:00
George Hotz
49dafe6d43 add gc tests [pr] (#9718) 
* add gc tests [pr]

* del

* more gc tests

* add NullGraph
 2025-04-03 14:08:32 +08:00
Ignacio Sica
bc91fffc5d fix gated store with index in python backend (#9703) 
* add default gate in index

* assert store

* add TestRendererFailures

- move test_gated_store_with_alu to new TestRenderFailures class for
tests that fail on multiple renderers
- add test_renderer_failures.py run on python CI

* add test for gated index in 2d

* test TestRenderFailures
 2025-04-03 12:48:28 +08:00
geohotstan
e1d7e47cca fix ONNX IsInf unintended dtype promotion (#9711) 
* add IsInf

* add corresponding test

* that float16 is kinda silly
 2025-04-02 22:46:15 -04:00
George Hotz
1714fc3ba4 start work on speed [pr] (#9707) 
* fix get_location

* fix get_location try 2

* clean up split_load_store [pr]

* SHR fixup [pr]
 2025-04-03 10:39:01 +08:00
George Hotz
0f1ffc2050 hotfix: cat tests 2048 instead of 256 2025-04-03 10:37:56 +08:00
Ignacio Sica
2d6d8b7355 add bf16 mfma support (#9695) 
* add bf16 mfma support

* skip tc if emulated_amd and dtypes is bf16

* hotfix
 2025-04-02 21:44:49 +08:00
chenyu
3b8d923692 remove skip LLVM in test_div_int (#9686) 2025-04-02 04:15:00 -04:00
George Hotz
e78e8722dc Revert "LDS noop and spec (#9669)" (#9691) 
This reverts commit 870b545ace.

Co-authored-by: Ignacio Sica <mignacio.sica@gmail.com>
 2025-04-02 15:31:32 +08:00
chenyu
c20f112e9f example test use z3 to verify valid simplification (#9684) 2025-04-02 01:05:52 -04:00
chenyu
bca0c85193 skip CI CPU test_data_parallel_resnet_train_step (#9685) 
flaky
 2025-04-02 01:04:54 -04:00
qazal
bb94f13e58 add RECORD_TRACEBACKS=1 option to process replay (#9679) 
* add RECORD_TRACEBACKS=1 option to process replay

* stack
 2025-04-02 11:58:27 +08:00
chenyu
c672716b38 improve vmin/vmax for IDIV (#9678) 2025-04-01 23:16:01 -04:00
chenyu
8dd88ad476 don't div_and_mod_folding for negative numerator with remainder (#9674) 
can be wrong in C div since it truncates towards zero
 2025-04-01 16:26:23 -04:00
chenyu
0e34f9082e helper functions for cstyle div mod [pr] (#9673) 2025-04-01 08:06:56 -04:00
Ignacio Sica
870b545ace LDS noop and spec (#9669) 
* init lds noop and lds_0 spec

* refactor lds helper test

* fix typo

* test all lds at the same time

* change comment

* comment

* start test_lds_full

* test_lds_tc

* add tc spec
 2025-04-01 18:44:55 +08:00
b1tg
d9af4cfc1b AMD_LLVM: tensor cores support (#9613) 
* tensor cores support

* test tesor cores codegen

* use rewrite rules

---------

Co-authored-by: b1tg <b1tg@users.noreply.github.com>
 2025-04-01 09:56:27 +08:00
Ignacio Sica
1444069c09 Uppercase K for dimension and lowercase k for kernel in linearizer tc helper test (#9649) 2025-03-31 19:05:36 +08:00
Ignacio Sica
baa67fd124 Uppercase N and M (standalone syntax change) (#9647) 2025-03-31 18:45:30 +08:00
Yvon Manzi
6652003839 Add cumprod to Tensor (#9629) 
* probably how cumprod should look like

* update _cumalu to work with MUL

* shorter

* cumprod testing

* clean

* more cleanup

* add cumprod to torch backend.

* make it look like cumsum

* mypy fix

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2025-03-30 21:49:18 -04:00