Gaurav Parmar
Gaurav Parmar
Thanks for pointing out this bug, it should be resolved now! There was a typo in the naming of the model.
Could you share what `accelerate` training config you were using when you encountered this error? -Gaurav
It looks like you are trying to do mixed precision (fp16) training. That might be the source of the issues.
Could you try with a config like this: ``` compute_environment: LOCAL_MACHINE debug: false distributed_type: MULTI_GPU downcast_bf16: 'no' gpu_ids: 0,1,2,3 machine_rank: 0 main_training_function: main mixed_precision: 'no' num_machines: 1 num_processes: 4 rdzv_backend:...
Ah I wrote the training code for fp32 training. It should be possible to change the training script to support fp16 training. I will take a look in a bit...
Hi, We used [this dataset](https://huggingface.co/datasets/wanng/midjourney-v5-202304-clean) for training our paired models. Additional details about the dataset are in our paper! -Gaurav
I just tried running these lines of code to test it: I was able to load the tokenizer without any errors. Could you perhaps try these two lines if you...
If you are training the unpaired model, the size of A and B folders do not need to be the same.
Based on this error, it looks like your image resolution is not a multiple of 8. What data-preprocessing are you using? -Gaurav
I think this is a problem that occurs because of mixed precision training. Could you try training without mixed precision?