Re:Replit - How to train your own LLM Models
by alpaca128 on Thursday April 20, 2023 @07:08AM
(#35638384)
Not everyone wants to depend on and trust a cloud service, and not everyone needs GPT-4 quality.
If there's a viable way to tune and run models locally they could still be useful if you don't need it to play chess and imitate a Python interpreter at the same time.
Re:Replit - How to train your own LLM Models
by ssivark on Thursday April 20, 2023 @09:05AM
(#35639319)
The original “pre-training” Is what’s expensive. The “fine-tuning” (also training that it modifies network weights) for instruction following or other tasks costs the thousand dollar range.
If there's a viable way to tune and run models locally they could still be useful if you don't need it to play chess and imitate a Python interpreter at the same time.