general_knowledge_model / generation_config.json
Nahush-27's picture
Push SFT GK model (MMLU 3k + NaturalReasoning 3k, LoRA r=64)
c7eae7a verified
raw
history blame
208 Bytes
{
"bos_token_id": 151643,
"do_sample": false,
"eos_token_id": [
151645,
151643
],
"max_new_tokens": 4096,
"pad_token_id": 151643,
"temperature": 1.0,
"transformers_version": "5.7.0"
}