Results tend to return summaries by default instead of raw transcription
#26
by
shaunck96
- opened
When using for transcription, the model tends to return summaries instead of raw text. Any recommendations to better align this behavior.
@shaunck96
Thank you for your interest in Phi-4-multimodal.
Can you share your prompt and audio files?
You need to use some special commands to get a transcription.Check online. I don't remember exactly what code to type but I know that I read it.
@shaunck96 You can take a look at the sample inference code https://huggingface.co/microsoft/Phi-4-multimodal-instruct/blob/main/sample_inference_phi4mm.py
Thanks a lot guys, will check this out.
nguyenbh
changed discussion status to
closed