Bot releases are hidden (Show)
fairscale
- v0.1.1
Published by msbaines almost 4 years ago
Fixed
- make sure pip package includes header files (#221)
fairscale
- v0.1.0
Published by msbaines almost 4 years ago
Added
- ShardedDataParallel with autoreduce (#157)
- cpu support for Pipe (#188)
- ShardedOptim: Distributed Grad Scaler (for torch AMP) (#182)
- OSS-aware clip grads, bridge sharded states (#167)
- oss: add rank_local_state_dict staticmethod (#174)
- support for PyTorch 1.7.0 (#171)
- Add implementation of AdaScale (#139)
Fixed
- pip package install (#196, #200)
fairscale
- v0.0.3
Published by msbaines almost 4 years ago
Added
- multi-process pipe (#90)
Fixed
- OSS+apex fix (#136)
- MegaTron+OSS DDP fix (#121)
fairscale
- v0.0.2
Published by msbaines almost 4 years ago
Added
- add ddp that works with oss with reduce() not all_reduce() (#19)
- support for PyTorch v1.6
- add mixed precision Adam (#40)
- Adam optimizer state scaling (#44)
Fixed
- properly restore a sharded optim state (#39)
- OSS restore state to proper device (#46)
- optim/oss: support optimizers with additional step kwargs (#53)
- optim/oss: fix state cast (#56)
- fix eval for oss_ddp (#55)
- optim/oss: work correctly with LRScheduler (#58)
fairscale
- v0.0.1
Published by msbaines about 4 years ago
- Initial release.
Package Rankings
Top 6.75% on Proxy.golang.org
Top 16.92% on Spack.io
Top 15.15% on Conda-forge.org
Badges
Extracted from project README
