That’s super cool! I think natural language processing should be put to even greater use going forward. Speech-to-text APIs are already available from various providers, so getting started with an API would be relatively easy and affordable. But it would be even better if you could pull it off without any monetary cost!
Login to reply
Replies (1)
Sooooon! I have a big Dell rack mount and a couple of headless GPUs ready to rock 'n' roll. Should load a 30 billion parameter model and under 30 seconds. Run multiple harnesses or agent workflows, local rag embedding waiting etc. shouldn't be an issue. Soon I will finally have a personal assistant!

