I need to use two different models in Spark and save them in the same folder. How can I deal with having both models with the same name pytorch_model.bin. If I rename one of them, how can I load the model?
Could you describe your use case a bit more and especially how Spark is used with PyTorch?
Also, could you explain what pytorch_model.bin
contains and how it’s created as it doesn’t seem to be a “standard” PyTorch state_dict
?