is model metadata/model id missing?
Hi, first some introduction: I don't speak Tunisian Arabic and this is the only transcription option I've found that seemingly works with this language. Thanks a lot for that.
Now the actual issue: I wanted to translate some TV shows where Tunisian Arabic is used via https://github.com/abdeladim-s/subsai. It supports hugging face models, but when selecting Hugging Face as a model, in model configuration, it wants to put in model_id and here is where I'm lost because nothing I tried worked. My suspicion is that this space needs some metadata files to be added which recognizes it as model and knows what to do with it. Would you be kind enough check it out?
Ok I've found out model id here would be SalahZa/Tunisian_Automatic_Speech_Recognition. Then I get another error, which is
SalahZa/Tunisian_Automatic_Speech_Recognition does not appear to have a file named config.json. Checkout 'https://huggingface.co/SalahZa/Tunisian_Automatic_Speech_Recognition/main' for available files.
Would it be possible for you to add config.json, similarly like the one at https://huggingface.co/openai/whisper-tiny/blob/main/config.json ?