You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In PocketFlow, we adopt multi-GPU training to speed-up this time-consuming training process. Our implementation is compatible with:
Horovod: a distributed training framework for TensorFlow, Keras, and PyTorch.
TF-Plus: an optimized framework for TensorFlow-based distributed training (only available within Tencent).
Is TF-Plus the private component developed by Tencent and not open-sourced yet? Are there any plan to open-source it?
The text was updated successfully, but these errors were encountered:
Is TF-Plus the private component developed by Tencent and not open-sourced yet? Are there any plan to open-source it?
The text was updated successfully, but these errors were encountered: