change "use_cache" to true to speed up decoding

Add chat_template from allenai/tulu-2-dpo-70b to tokenizer_config.json (#4)

Update README.md

Adding `safetensors` variant of this model (#1)

first commit

initial commit