David Hou 4b95350c41 fp16 resnet (without expand backwards sum in float, doesn't work) (#3816) ...

* fp16 resnet

* cast running mean and var back to default float

* extra cast

* check symbolic no overflow

* add linearizer failure

* loss scaler after grad contig

* oops

* i think this works

* don't loss scale fp32

* remove overflow test case

* remove symbolic bounds check

* loss scaler should be float

* temporarily disable padto cuz bug

shruggie

* make running stats in batchnorm float32?

* calculate lars stuff in fp32?

* oops

* remove most changes

* move loss scaler out of optimizer

* no more FP16 var

* oops

---------

Co-authored-by: chenyu <chenyu@fastmail.com>

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

..

bert.py

multilazybuffer fix (#3609)

2024-03-04 17:36:23 -05:00

convnext.py

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00

efficientnet.py

simple LoadOps.ASSIGN (#3745)

2024-03-14 20:44:34 -07:00

llama.py

restrict assignment to base (#3809)

2024-03-18 15:33:06 -07:00

mask_rcnn.py

fix example that calls Tensor.__bool__ (#3650)

2024-03-07 16:59:26 -05:00

resnet.py

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

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

retinanet.py

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00

rnnt.py

create engine folder and move code (#3948)

2024-03-26 20:38:03 -07:00

transformer.py

replace with tensor op (#3099)

2024-01-12 14:13:40 -05:00

unet3d.py

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00

vit.py

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00