Update?
#4
by
blankreg
- opened
I noticed tokenizer_config and generation_config were updated for both 7B and 14B after this gguf release, does it need updating? Sorry for the noob question
No worries, no those specific changes don't affect llama.cpp
They updated model_max_length which doesn't matter because GGUF uses max_position_embeddings first then falls back to model_max_length if needed
The other changes were only relevant for other tools
Thanks for your work!
blankreg
changed discussion status to
closed