Fixed some questions. Also an additional prompt for the LoRA models to deal with the loss of attention with LoRA models, and be better aligned with the prompt used while training.
This is a temporary solution until our LoRA models are trained with the same prompt as our current prompt_cot.md
Fixed some questions. Also an additional prompt for the LoRA models to deal with the loss of attention with LoRA models, and be better aligned with the prompt used while training.
This is a temporary solution until our LoRA models are trained with the same prompt as our current
prompt_cot.md