phi4-mm-gptq / generation_config.json
Swicked86's picture
Add W4A16 GPTQ quantization with speech/vision LoRA adapters
7434286 verified
Raw
History Blame Contribute Delete
169 Bytes
{
"_from_model_config": true,
"bos_token_id": 199999,
"eos_token_id": [
200020,
199999
],
"pad_token_id": 199999,
"transformers_version": "4.57.6"
}