Zuolei Li

How to train the network with checkpoint file?

VisionLanguageFusionModule

1

I find in this module you use multiplication instead of addition or concatenation. ``` tgt2 = self.multihead_attn(query=self.with_pos_embed(tgt, query_pos), key=self.with_pos_embed(memory, pos), value=memory, attn_mask=None, key_padding_mask=memory_key_padding_mask)[0] tgt = tgt * tgt2 return tgt...

Test Time Adaptation

2

I found that the project uses the test time augmentation instead of test time adaptation. Where is the test time adaptation used?

Bidirectional Matching

3

In forward matching, what is the dimesion of the points on the reference mask Pr? And what does the L means?

The loss for weakly supervised learning?

1

I can not find the code about the loss for weakly supervised learning. Could you tell me how to find them?

Thanks for your great work! Can you update this file? Now the following code has fault because ```from_preloaded``` is removed: ``` lerobot_dataset = LeRobotDataset.from_preloaded( repo_id=repo_id, hf_dataset=hf_dataset, episode_data_index=episode_data_index, info=info, videos_dir=videos_dir, )...