Sean Naren bacabaebaf Sharded Accelerator 1/n: Expose clip gradients to plugins via abstract class (#4639) ... * Added abstract precision plugin to expose clip_gradients function, use within accelerator to clip gradients * Exclude model from override, keep optimizer (needed for sharded clip gradients), add override for O2 support apex * Fix doc * Applied codereview changes * Refactored clip function to encapsulate tpu changes with tpu accelerator. Default to standard clip function for vanilla torch * Pass correct grad clip val * Moved var to property * Apply code review suggestions		2020-11-12 17:18:09 +00:00
..
__init__.py	ref: apex plugin (#3502)	2020-09-15 06:02:42 -04:00
apex.py	Sharded Accelerator 1/n: Expose clip gradients to plugins via abstract class (#4639)	2020-11-12 17:18:09 +00:00
ddp_plugin.py	Enable DDP Plugin to pass through args to LightningDistributedDataParallel (#4382)	2020-10-27 12:27:59 +00:00
native_amp.py	Sharded Accelerator 1/n: Expose clip gradients to plugins via abstract class (#4639)	2020-11-12 17:18:09 +00:00
plugin_connector.py	Enable custom apex and amp plugins (#4355)	2020-10-25 17:11:07 -04:00
precision_plugin.py	Sharded Accelerator 1/n: Expose clip gradients to plugins via abstract class (#4639)	2020-11-12 17:18:09 +00:00