Question about inference_hf.py

declare-lab / tango

A family of diffusion models for text-to-audio generation.

https://tango2-web.github.io/

Other

991 stars 79 forks source link

Question about inference_hf.py #34

Open RookieJunChen opened 12 months ago

RookieJunChen commented 12 months ago

I noticed that you updated the inference_hf.py file in your tango repository. May I kindly ask how it differs from the inference.py file?

deepanwayx commented 12 months ago

The inferece_hf.py will compute the evaluation metrics (reported in the paper) from the checkpoints we uploaded in huggingface. For each text prompt, only one audio sample will be generated.

The inferece.py has additional functionality to generate multiple audio samples for each text prompt and then re-ranking them according to CLAP scores.