Mistral-7B-v0.1-GPTQ-8bit-g64 / quantize_config.json
iproskurina's picture
AutoGPTQ model for mistralai/Mistral-7B-v0.1: 8bits, gr64, desc_act=False
e3a5cbc verified
raw
history blame contribute delete
210 Bytes
{
"bits": 8,
"group_size": 64,
"damp_percent": 0.01,
"desc_act": false,
"static_groups": false,
"sym": true,
"true_sequential": true,
"model_name_or_path": null,
"model_file_base_name": null
}