github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-06 11:32:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

Support TP-compatible Torch AMP and Update trainer API (#27)

Browse Source 
* Add gradient accumulation, fix lr scheduler

* fix FP16 optimizer and adapted torch amp with tensor parallel (#18)

* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes

* fixed trainer

* Revert "fixed trainer"

This reverts commit 2e0b0b7699.

* improved consistency between trainer, engine and schedule (#23)

Co-authored-by: 1SAA <c2h214748@gmail.com>

Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
This commit is contained in:
Frank Lee
2021-11-18 19:45:06 +08:00
committed by GitHub
parent 2b05de4c64
commit 3defa32aee
80 changed files with 2194 additions and 1584 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
colossalai/nn/lr_scheduler/linear.py
View File

@@ -28,18 +28,3 @@ class LinearWarmupLR(_LRScheduler):
else:
return [(self.total_steps - self.last_epoch) / (self.total_steps - self.warmup_steps) * lr for lr in
self.base_lrs]
@LR_SCHEDULERS.register_module
class LinearWarmupDecay(_LRScheduler):
def __init__(self, optimizer, total_steps: int, warmup_steps: int = 0, last_epoch: int = -1, **kwargs):
self.warmup_steps = int(warmup_steps)
self.total_steps = total_steps
super().__init__(optimizer, last_epoch=last_epoch)
def get_lr(self):
if self.last_epoch < self.warmup_steps:
return [(self.last_epoch + 1) / self.warmup_steps * lr for lr in self.base_lrs]
else:
return [(self.total_steps - self.last_epoch - 1) / (self.total_steps - self.warmup_steps) * lr for lr in
self.base_lrs]