I have trained a model for caption generation , I've used VIT as an encoder and GPT2 as a decoder. After training, these files were saved: model.pth and image-caption-generator checkpoint-4045 : config.json, generation_config.json, model.safetensors, optimizer.pt, preprocessor_config.json, rng_state.pth, scheduler.pt, trainer_state.json, training_args.bin
I've saved the outputs to my computer then opened the notebook and upload them but I didn't know how to load them again to make predictions