Why the AGD model does not exist in the optimizer

ygjwd12345 / TransDepth

Code for Transformers Solve Limited Receptive Field for Monocular Depth Prediction

MIT License

172 stars 20 forks source link

Closed 760677482 closed 2 years ago

760677482 commented 2 years ago

The Training parameters are only encoder and decoder

ygjwd12345 commented 2 years ago

sorry for this stupid error. I will modify it immediately.

ygjwd12345 commented 2 years ago

done

760677482 commented 2 years ago

done

Thanks for your reply

760677482 commented 2 years ago

Should the AGD module also need be initialized before training like the decoder？

ygjwd12345 commented 2 years ago

We don't try this.