tinygrad

geohot/tinygrad

Fork 0

forked from tinygrad/tinygrad

Files

T

History

Elias WahlandGitHub 7bfa9101c0 Float in scaled dot product attention (#4985 )

* Monkeypatch scaled-dot-product-attention

* Use dot instead of matmul

* new api

* imports

* least_upper_dtype

2024-06-18 08:16:41 -04:00

training_submission_v4.0/tinycorp

feat: more mlperf fixes (#4505 )

2024-05-09 20:50:20 -07:00

dataloader.py

Multiprocessing UNet3D dataloader (#4801 )

2024-06-02 11:30:47 -04:00

helpers.py

Disable dropout (#4837 )

2024-06-04 18:57:26 -04:00

initializers.py

Float in scaled dot product attention (#4985 )

2024-06-18 08:16:41 -04:00

losses.py

[MLPerf][UNet3D] Add DICE loss + metrics (#4204 )

2024-04-17 20:09:33 -04:00

lr_schedulers.py

fp16 resnet (without expand backwards sum in float, doesn't work) (#3816 )

2024-03-28 01:25:37 -04:00

metrics.py

[MLPerf][UNet3D] Add DICE loss + metrics (#4204 )

2024-04-17 20:09:33 -04:00

model_eval.py

add mlperf train subset of openimages (#4841 )

2024-06-05 10:10:11 -04:00

model_spec.py

move globalcounters to ops (#2960 )

2024-01-01 14:21:02 -08:00

model_train.py

Float in scaled dot product attention (#4985 )

2024-06-18 08:16:41 -04:00

README

start on mlperf models

2023-05-10 16:30:49 -07:00

README

Each model should be a clean single file.
They are imported from the top level `models` directory

It should be capable of loading weights from the reference imp.

We will focus on these 5 models:

# Resnet50-v1.5 (classic) -- 8.2 GOPS/input
# Retinanet
# 3D UNET (upconvs)
# RNNT
# BERT-large (transformer)

They are used in both the training and inference benchmark:
https://mlcommons.org/en/training-normal-21/
https://mlcommons.org/en/inference-edge-30/
And we will submit to both.

NOTE: we are Edge since we don't have ECC RAM