cosail-knu
/

llama2-2.7b_8

8-bit precision

Model card Files Files and versions

llama2-2.7b_8 / generation_config.json

knu-mhan's picture

Add 8-bit HQQ quantized Sheared-LLaMA-2.7B model

79ff877 about 2 months ago

history blame contribute delete

132 Bytes

	{
	"_from_model_config": true,
	"bos_token_id": 1,
	"eos_token_id": 2,
	"pad_token_id": 0,
	"transformers_version": "4.49.0"
	}