llama3-s-instruct-v0.2/generation_config.json

{
  "_from_model_config": true,
  "bos_token_id": 128000,
  "eos_token_id": [
    128001,
    128008,
    128009
  ],
  "transformers_version": "4.44.0"
}