Upload LlamaForCausalLM
Upload tokenizer
Add llama3-8B-EIP-8bit-lora model trained on DGX-1
initial commit
