Orion-14B-Chat-Int4 / quant_config.json
DachengZhang's picture
init
9c7e7ed
raw
history blame
126 Bytes
{
"zero_point": true,
"q_group_size": 128,
"w_bit": 4,
"version": "GEMM",
"modules_to_not_convert": null
}