Open AlanWang0o0 opened 10 months ago
Hi, the links to data and checkpoints are back online now. In terms of the code implementation of TGE, we are planning to release them along with an advanced version of GraspGPT in the near future.
Chao
So I can't get the results posted in the manuscript? The TGE module is only two cross-attention layer according to your paper, seems easy to be implemented. I wonder why I can't achieve the performance after I followed all the implementation details.
Any more follow-ups on this issue? I've met the same problem :(
I re-implement the Task-Oriented Grasp Evaluator (TGE) module proposed in your paper, following the architecture depicted in your letter. Howerer, the performance is even worse. I am sure that I follow all the details (e.g. the dimension, the multi-head attention). I wonder if there are any more information about the TGE module, and what should I do to achieve the results you provided in the paper.
By the way, the links for downloading the LA-TaskGrasp and the checkpoints are invalid.