github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-04-29 03:00:14 -04:00
Code Issues Packages Projects Releases Wiki Activity
12,053 Commits 248 Branches 14 Tags
b47397ab1796fa4f8facedcd4575c80a685cfa5b
Commit Graph

1261 Commits

Author SHA1 Message Date
chenyu
d57d24c7d4 Buffer.as_buffer -> Buffer.as_memoryview [pr] (#14535) 
it casts to memoryview. also inline the as_typed_buffer checks to Tensor._data
 2026-02-04 11:31:11 -05:00
George Hotz
d59e6e7a37 move more tests to test/null, split some existing ones (#14512) 
* move more tests to test/null, split some existing ones

* null work

* null work

* move more

* fixes

* move PIL

* PIL in CLIP

* don't move that
 2026-02-03 20:20:20 +08:00
George Hotz
dd2de4f838 rename all DEFINE_GLOBAL to PARAM (#14511) 2026-02-03 15:09:38 +08:00
wozeparrot
bbcd3d67a3 fa: faster (#14453) 2026-02-02 21:34:17 -08:00
qazal
616e9c1483 CDNA assembly gemm in tensor.py with flag (#14310) 
* work

* work

* the assembly

* remove the old one

* remove ws bufs, assert splitk

* notes cleanup

* work

* gemm args

* gemm in mixins would be nice

* add gemm gradient

* print counters

* the realize is for DEBUG=2 aesthetics

* dedup

* rewrite to python dsl, no list copies

* leave that

* add B, M, N, K to gemm name

* it's M0 not NULL

* fp16 support

* test cleanup + more gemms

* work from viz

* more work

* gemm batch_size

* xccg path work

* tiny comments on the label naming

* s_waitcnt
 2026-01-31 22:34:14 +09:00
George Hotz
c9a3ddb341 benchmark llama walltime script (#14454) 
* benchmark llama walltime script

* adj layers
 2026-01-31 10:21:54 +08:00
George Hotz
f5346d6a1a fix USE_ATOMICS for non float dtypes and make it the default (#14444) 
* embedded multistep test

* complex test

* with jit

* fix dtypes and reenable USE_ATOMICS

* that test didn't catch anything
 2026-01-31 09:44:16 +08:00
George Hotz
ee2c78709d mlperf/llama: disable USE_ATOMICS for now 2026-01-31 00:42:08 +08:00
George Hotz
838cd078bc use atomics for embedding backward (#14400) 
* embedding is slow

* failing

* float is fine

* null

* it fails

* simplify embedding with broadcasting

* ATOMIC_ADD incoming

* min change

* simpler test

* better test

* fix test

* real test

* simpler

* cleanups

* types and names

* _zero_kernel

* grad multi

* hack

* none

* multi unshard

* more for call

* don't tag in call

* good

* call_multi

* call_multi wow claude is useless

* embedding backward mutli test

* test passes

* fix as_param

* shape_to_shape_arg

* add clip

* before cast

* fix spec=2, use atomics
 2026-01-30 18:10:59 +08:00
George Hotz
793afbd473 simplify nn.Embedding, support AFTER in CUSTOM_KERNEL (#14419) 2026-01-29 17:22:13 +08:00
wozeparrot
4845e42135 llama3 gradacc fixes (#14414) 2026-01-28 19:12:39 -08:00
nimlgen
aec1ae0de1 llama: set manual_seed (#14409) 2026-01-28 14:40:00 -08:00
George Hotz
0c6b3f50aa add marker to llama training (#14401) 2026-01-28 22:44:28 +08:00
Jakob Sachs
2b7c00d3d2 fix sd-example dtype for CLIP embeddings (#14397) 2026-01-28 09:07:19 -05:00
qazal
5bffa17f82 llama train: better NULL=1 EMULATE=AMD_CDNA4 dev experience (#14395) 
* beam opens devices

* switch to hip renderer

* amd: true?

* llvm true is for test_autogen
 2026-01-28 17:31:22 +09:00
wozeparrot
e496547720 llama3 gradacc (#14291) 2026-01-27 19:48:10 -08:00
chenyu
db010a31be IGNORE_OOB -> CHECK_OOB [pr] (#14374) 
flip the meaning
 2026-01-27 12:20:59 -05:00
wozeparrot
a987a4abc3 feat: llama8b dev_beam.sh (#14358) 2026-01-26 14:51:23 -08:00
nimlgen
e152f1b0f5 llama: use ALL2ALL (#14353) 2026-01-26 22:01:53 +03:00
George Hotz
11ce1e847d llama train: null device support 2026-01-26 08:53:05 +08:00
wozeparrot
963c59ebdb fix: pull fixes from gradacc branch (#14296) 2026-01-22 23:07:54 -08:00
George Hotz
52b989c6c8 don't place consts early + fixes from anthropic challenge (#14286) 
* don't place consts early

* add anthropic challenge

* with ref

* do we still have to devectorize bools?

* tests pass

* just WHERE

* fine, revert that

* fine, revert

* only index

* z3 validator doesn't support vectorized

* Revert "z3 validator doesn't support vectorized"

This reverts commit 1b7930ecb3.

* z3 not for vec

* no spec

* VLIWRenderer

* loop unrolling

* better comments

* cleanups

* skip cast

* renderer

* cleanups

* prints

* no hack

* hacks

* bump to 11

* reg warning

* lil clean

* cleaner renderer
 2026-01-23 10:48:39 +09:00
wozeparrot
c1d14ea832 llama8b train fixes (#14264) 2026-01-20 20:34:47 -08:00
wozeparrot
ba90e1b52e feat: script to run llama8b training (#14239) 2026-01-20 12:44:06 -08:00
C T
26f8b12e01 Whisper audio helpers (mel filters in tinygrad) (#13478) 
* add whisper audio helpers for stft/mel/resample

* cleanup

* add whisper stft test

* make only stft test explicitly depend on librosa

* extract sinc_window_kernel

* dehardcode device

* use same device argument

* simplify

* type annotate

* ruff format audio_helpers.py

* ruff format test_whisper.py

* add WHISPER_NEW_STFT

* rename

* undo ruff format changes

* use new stft and mel for whisper

* remove stft test that depends on librosa

* remove whitespace

* add Tensor.log10 with test\test_ops.py::TestOps::test_log10

* use Tensor.log10

* fix lint

* future: remove unused STFT class

* future: remove resample code since it isn't used (yet)

* match openai with pad_mode="reflect"

* pad_to

* future: cut resample leftovers

* cleanup

* add mel tests

* future: cut stft

* future: cut non-mel prep_audio changes

* reduce diff

* move audio_helpers.py to examples

* reduce whitespace

* fix imports

* reduce whitespace

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2026-01-20 10:50:02 -05:00
wozeparrot
a879b54234 tk: fa jit fix (#14170) 2026-01-16 16:38:45 -08:00
b1tg
0fbc551622 train bert with fp8 (#13874) 
* fp8 train

* clean

* lint

* test fix from #13439

* skip first/last layer

* rm __init__, restore unroll <=32 check

* tests

* clean test, remove unused

* multi-gpu test, clean quantize_to_fp8

* remove bert contiguous

* run script

* test: better check

* run script search

* add seed in bert data shuffle

* move script to mi350x folder

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2026-01-09 09:21:59 -05:00
b1tg
241f0402b4 add seed in bert data shuffle (#14054) 2026-01-07 10:02:05 -05:00
chenyu
87f4bc5446 update variable names around jit [pr] (#14049) 
lbs, st_vars_dtype_device and rawbuffers no more
 2026-01-06 22:32:41 -05:00
Francis Lata
fac137779e remove flux1 seed image (#13843) 2025-12-27 00:45:11 -05:00
chenyu
da1cb6a9ec update llama dataloader (#13825) 
separate creating dataset from itererating over the dataset to not create eval data for each eval
 2025-12-24 17:42:08 -05:00
chenyu
903753c60c llama wandb logging (#13822) 2025-12-24 10:24:59 -05:00
chenyu
27d899ce97 TRAIN=0 to only eval llama (#13804) 2025-12-22 11:55:46 -05:00
chenyu
39d962106f update llama logging (#13803) 
```
REWRITE_STACK_LIMIT=1000000 SMALL=1 BASEDIR=/raid/datasets/c4-8b SAMPLES=1000 BS=8 DP=8 DEFAULT_FLOAT=bfloat16 OPTIM_DTYPE=bfloat16 LLAMA3_SIZE=8B SEQLEN=1024 PYTHONPATH=. MODEL=llama3 python3 examples/mlperf/model_train.py

    1 93.44 s run, 11.8750 loss, 0.000000000001 LR, 642.43 GB used,  19644.30 GFLOPS
    2 101.78 s run, 11.8750 loss, 0.000000000001 LR, 1454.57 GB used,  17039.35 GFLOPS
    3 7.34 s run, 11.8750 loss, 0.000000000002 LR, 1454.57 GB used, 236258.78 GFLOPS
    4 4.32 s run, 11.8750 loss, 0.000000000002 LR, 1454.57 GB used, 401488.40 GFLOPS
    5 4.36 s run, 11.9375 loss, 0.000000000003 LR, 1454.57 GB used, 398116.13 GFLOPS
    6 4.32 s run, 11.8750 loss, 0.000000000003 LR, 1454.57 GB used, 401878.60 GFLOPS
    7 4.34 s run, 11.8750 loss, 0.000000000004 LR, 1454.57 GB used, 399822.57 GFLOPS
    8 4.35 s run, 11.8750 loss, 0.000000000004 LR, 1454.57 GB used, 398512.24 GFLOPS
    9 4.36 s run, 11.8750 loss, 0.000000000005 LR, 1454.57 GB used, 397832.61 GFLOPS
   10 4.40 s run, 11.8750 loss, 0.000000000005 LR, 1454.57 GB used, 394520.83 GFLOPS
```
 2025-12-22 11:28:29 -05:00
George Hotz
45c459848d remove more stale stuff (#13765) 
* remove more stale stuff

* remove disassemblers/adreno

* stale
 2025-12-19 17:14:56 -04:00
George Hotz
df6cde8a00 cleanup stale examples/extra (#13764) 
* cleanup stale files

* examples

* move those back

* old

* delete more
 2025-12-19 16:27:37 -04:00
chenyu
7cd7593c5d add script to train bert on mi350x (#13743) 
adapted from mi300 config
 2025-12-17 16:54:04 -05:00
chenyu
e428fbfab6 verify dtype of llama model params (#13719) 2025-12-16 12:32:02 -05:00
chenyu
6cad622f59 don't FREE_INTERMEDIATE in bert (#13684) 
hangs green hcq consistently after an hour of training
 2025-12-14 14:27:42 -05:00
chenyu
fcaed1e1dd don't use empty in bert fake data (#13661) 
somehow jit does not count empty as input
 2025-12-12 15:59:50 -05:00
chenyu
01e9ad0d52 clean up bert next_data (#13650) 
train iter was designed to never stop for both real and fake data
 2025-12-11 22:56:28 -05:00
chenyu
5034c6fb37 reenable FREE_INTERMEDIATE for bert (#13639) 
* reenable FREE_INTERMEDIATE for bert

* comment
 2025-12-10 12:08:09 -05:00
chenyu
016a59cafa remove contiguous and use where in EmbeddingBert (#13632) 2025-12-09 15:49:21 -05:00
chenyu
2471b49e45 minor bert / llama change from grad acc branch (#13622) 
* minor bert / llama change from grad acc branch

* revert those
 2025-12-08 16:04:14 -05:00
chenyu
b981b6f89e remove old llama grad_acc (#13611) 
* remove old llama grad_acc

* GRADIENT_ACC_STEPS=1
 2025-12-07 13:03:47 -05:00
chenyu
4562f217e1 more bert updates (#13597) 
prep split jit
also lower BS to 72
 2025-12-06 08:32:43 -05:00
chenyu
cb4c6324ef revert bert grad accumulation (#13596) 
prep for the new split jit style
 2025-12-05 17:30:08 -05:00
chenyu
89f9e1dcd5 add SGD to beautiful_mnist (#13571) 2025-12-04 12:17:29 -05:00
George Hotz
96d16675fe update examples/gradaccum_mnist.py to use the JIT 2025-12-03 16:11:42 -08:00
George Hotz
a4c4e48385 add LUNIQUE op (#13554) 2025-12-03 14:34:34 -08:00