Edit model card

This model is a base llama-3-8b model but set up for chat finetuning by modifying the tokenizer chat-tempalte and setting up special tokens for chat training.

Downloads last month
78
Safetensors
Model size
4.65B params
Tensor type
F32
BF16
U8
Inference API
Input a message to start chatting with pookie3000/llama-3-8b-bnb-4bit-for-chat-training.
This model can be loaded on Inference API (serverless).