2 ms·
This is super cool. Do you know if any of the inference backends (llama.cpp, vllm, etc) support this technique?
by meatmanek 4mo ago
This is super cool. Do you know if any of the inference backends (llama.cpp, vllm, etc) support this technique?
- iaw 4mo agovLLM supports "banning" certain tokens but I don't know if it can dynamically reduce them. To my knowledge you can also "ban" with llama.cpp but it is passed in the API call rather than to the server at initialization.