Can't specify model path.
I've downloaded the model to a mnt folder, since many users could use this model, however I can't pass the model location to read from there. It starts downloading the model again in local cache directory. Is there a workaround to this? I've always done this using open source models.
` model_id = "/mnt/AI-Models/huggingface/ResembleAI/chatterbox" ChatterboxMultilingualTTS.from_pretrained(model_id=model_id, device=device)
tts = TTSGenerator(model_id=model_id) # create an instance
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ahammad/AI-Interviewer/src/text_to_speech.py", line 12, in init
self.model = ChatterboxMultilingualTTS.from_pretrained(model_id=model_id, device=device)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: ChatterboxMultilingualTTS.from_pretrained() got an unexpected keyword argument 'model_id'
`
You can use ChatterboxMultilingualTTS.from_local method:
multilingual_model = ChatterboxMultilingualTTS.from_local(
device=device, ckpt_dir="model")