search

cyclomon / UNSB

Official Repository of "Unpaired Image-to-Image Translation via Neural Schrödinger Bridge" (ICLR 2024)
MIT License
154 stars 5 forks source link

Training with batch size > 1 #13

Closed juanprietob closed 7 months ago

juanprietob commented 7 months ago

I have a large dataset. Is it possible to train with a batch size larger than 1? This fails if the batch size is larger than 1 https://github.com/cyclomon/UNSB/blob/main/models/sb_model.py#L178

cyclomon commented 7 months ago

Hi, I corrected the code for multi-batch training.

The training is now available for multi-batch, but I did not evaluated the training output in multi-batch case. There might be some difference compared to single-batch case.