William FalconandJirka 10b16dbfab made ddp the default if no backend specified with multiple GPUs (#1789) ...

* made ddp the default if no backend specified with multiple GPUs

* fix

* spawn

Co-authored-by: Jirka <jirka.borovec@seznam.cz>

2020-05-12 06:54:23 -04:00

..

__init__.py

Feature: auto scale batch size (#1638)

2020-05-09 08:28:36 -04:00

auto_mix_precision.py

support for native amp (#1561)

2020-04-23 14:47:08 -04:00

callback_config.py

Fix typo (#1750)

2020-05-07 09:25:54 -04:00

callback_hook.py

Progress bar callback (#1450)

2020-04-23 20:46:18 -04:00

data_loading.py

added warning for None dataloader (#1745)

2020-05-07 09:26:41 -04:00

deprecated_api.py

Progress bar callback (#1450)

2020-04-23 20:46:18 -04:00

distrib_data_parallel.py

made ddp the default if no backend specified with multiple GPUs (#1789)

2020-05-12 06:54:23 -04:00

distrib_parts.py

Join Horovod workers at the end of trainer.fit() to prevent race conditions following training (#1786)

2020-05-12 09:15:25 +00:00

evaluation_loop.py

enable fast_dev_run without a validation loop (#1779)

2020-05-11 11:30:22 -04:00

ignored_warnings.py

Dim 0 warning (#256)

2019-09-26 13:20:54 -04:00

logging.py

[WIP] Reduction when batch size < num gpus (#1609)

2020-05-02 11:01:44 -04:00

lr_finder.py

Feature: auto scale batch size (#1638)

2020-05-09 08:28:36 -04:00

model_hooks.py

Fix typo (#1750)

2020-05-07 09:25:54 -04:00

optimizers.py

dataloaders with fast_dev_run (#1787)

2020-05-11 23:32:44 -04:00

supporters.py

Fixed configure optimizer from dict without "scheduler" key (#1443)

2020-04-10 11:43:06 -04:00

trainer.py

fixed native amp + ddp (#1788)

2020-05-12 00:25:06 -04:00

training_io.py

Fix saving native AMP scaler state (#1777)

2020-05-11 21:38:37 -04:00

training_loop.py

Also update progress_bar in training_epoch_end (#1724)

2020-05-08 23:31:56 -04:00

training_tricks.py

dataloaders with fast_dev_run (#1787)

2020-05-11 23:32:44 -04:00