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 the repository's documentation, there are many references to pretrained weights and checkpoints.
However, I can't find any checkpoints in the code or any links to a Hugging Face repo where they might be stored. Does such a thing exist, or should I try to create my own?
The text was updated successfully, but these errors were encountered:
If you are referring to the the resume_path argument in the training script here: this is an extra functionality used to resume training from a checkpoint, and is not necessary if you want to train from scratch on BridgeData and/or SOAR data.
If you are referring to the checkpoint_path when deploying the policy on robots here: you should pass in the path to your trained checkpoint (which you can train from model_training/experiments/train.py). Currently we don't have model weights uploaded to Hugging Face yet. But let me know if you have a specific use case and we can release related checkpoints.
In the repository's documentation, there are many references to pretrained weights and checkpoints.
However, I can't find any checkpoints in the code or any links to a Hugging Face repo where they might be stored. Does such a thing exist, or should I try to create my own?
The text was updated successfully, but these errors were encountered: