Closed mo230761 closed 1 month ago
Total resources cost is very close to SD1.5. Specifically, we use 120 GB VRAM (virtual memory) during training.
Thank you very much for your answer. If I set the batch size to 1 and the precision to 16, can I train on a 32G V100?
I believe that would working well. Please feel free to re-open the thread if you encounter any other issues.
I would like to know the minimum VRAM required to train the StyleBooth model. Could you tell me? Thank you!