Whisper like ASR model but with some advanced ideas. Experimental. Full script just install dependencies and run. The model included is -not- trained. Its a blank (tabula rasa) newly intialized version of the script "medium" sized. I'm experimenting with some of the new stuff from the vision llm people but with audio.. Here is a super cool paper: https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.949142/full
Updated. Was having some issues there with the hybrid attention and tensor sharing.. fixed.!
- Downloads last month
- 12