Error when using a custom model

Hello, I am currently trying to use a different `faster_whisper_custom_model_path` model such as `https://huggingface.co/primeline/whisper-tiny-german-1224` but when using it with `run_server.py`, the script raises a ValueError at whisper_live/server.py:416 ``` if faster_whisper_custom_model_path is not None and not os.path.exists(faster_whisper_custom_model_path): raise ValueError(f"Custom faster_whisper model '{faster_whisper_custom_model_path}' is not a valid path.") ``` Command for running the service: ``` python3 run_server.py --port 9090 \ --backend faster_whisper \ --max_clients 4 \ --max_connection_time 600 \ -fw primeline/whisper-tiny-german-1224 ``` but if I understoof the code correctly, this string is later used to pull the model from the huggingface hub. Is there a reason, why this string must be an existing path? Also, when providing a valid path with the downloaded weights and tokenizer inside, a 404 error is raised at whisper_live/backend/faster_whisper_backend.py:140 ``` local_snapshot = snapshot_download( repo_id = model_ref, repo_type = "model", ) ``` when running the following script: ``` from whisper_live.client import TranscriptionClient client = TranscriptionClient( "localhost", 9090, lang="de", translate=False, model="tiny", # also support hf_model => `Systran/faster-whisper-small` use_vad=True, save_output_recording=True, # Only used for microphone input, False by Default output_recording_filename="./output_recording.wav", # Only used for microphone input mute_audio_playback=False, # Only used for file input, False by Default enable_translation=False, target_language="de", ) client() ``` Does it make sense to ignore this error and download the model (tried and works) or does this break something?