How can the trained model be exported to .pt format?
Could this model be converted to the .pt format used in https://github.com/openai/whisper/? I want to run them with https://github.com/ggerganov/whisper.cpp
Does this answer your question? https://github.com/openai/whisper/discussions/64#discussioncomment-4117291
It seems lots of names have changed in the model.
https://github.com/ggerganov/whisper.cpp/issues/157#issuecomment-1320915269
Hey @jzju - it looks like the issue you've linked is related to the https://github.com/ggerganov/whisper.cpp repo rather than a HF one! We can help with exporting the model to a PyTorch state dict (as per my previous answer) and HF model related questions, but this is outside the scope of what we know / can assist with!
The repo https://github.com/ggerganov/whisper.cpp is the best place to ask how to bring a PyTorch state dict into the right format for the whisper.cpp toolkit
how can i save and restore the training checkpoints when the connection is lost in google colab? colab disconnect run before i finish my training ...i want to save my training and i resume it in another colab session
Hey @Martha-987 , I've opened a new discussion for your question: https://huggingface.co/sanchit-gandhi/whisper-small-hi/discussions/3