Glenn Jocher 9b4e05439c

Fix AMP check tolerance (#7937) ...

Adjust to 5%, fixes failing Colab AMP check with V100 (1.5% different) with 200% safety margin.

2022-05-23 15:57:16 +02:00

..

aws

…

docker

Update Dockerfile-arm64 (#7860)

2022-05-17 16:35:12 +02:00

flask_rest_api

Code refactor (#7923)

2022-05-22 15:24:08 +02:00

google_app_engine

…

loggers

Refactor collections and fstrings (#7821)

2022-05-15 16:38:26 +02:00

__init__.py

Update bytes to GB with bitshift (#6886)

2022-03-07 19:26:37 +01:00

activations.py

[pre-commit.ci] pre-commit suggestions (#7279)

2022-04-04 22:47:00 +02:00

augmentations.py

precommit: yapf (#5494)

2022-03-31 16:52:34 +02:00

autoanchor.py

Refactor collections and fstrings (#7821)

2022-05-15 16:38:26 +02:00

autobatch.py

Add PyTorch AMP check (#7917)

2022-05-22 13:41:18 +02:00

benchmarks.py

Add check_yaml() to benchmarks.py (#7916)

2022-05-21 19:39:19 +02:00

callbacks.py

Simplify callbacks.py return (#7333)

2022-04-07 16:15:01 +02:00

dataloaders.py

cv2.resize interpolation fix (#7903)

2022-05-20 12:59:05 +02:00

downloads.py

Pin downloads to release version (#7790)

2022-05-13 00:24:53 +02:00

general.py

Fix AMP check tolerance (#7937)

2022-05-23 15:57:16 +02:00

loss.py

Swap unsafe_chunk() for chunk() (#7362)

2022-04-10 13:53:28 +02:00

metrics.py

Bug fix mAP0.5-0.95 (#6787)

2022-05-20 11:33:10 +02:00

plots.py

Add @threaded decorator (#7813)

2022-05-14 16:12:08 +02:00

torch_utils.py

Added Windows cmd to count GPU devices (#7891)

2022-05-19 15:48:44 +02:00