multilingual-e5-large-quantized / quantize_config.json
freefallr's picture
upload quantized model
fdee064
{
"per_channel": true,
"reduce_range": true,
"per_model_config": {
"model": {
"op_types": [
"Reshape",
"ConstantOfShape",
"Shape",
"Div",
"Slice",
"Transpose",
"ReduceMean",
"Pow",
"Equal",
"Unsqueeze",
"Mul",
"Concat",
"MatMul",
"Add",
"Softmax",
"Constant",
"Sqrt",
"Erf",
"Where",
"Expand",
"CumSum",
"Gather",
"Sub",
"Not",
"Cast"
],
"weight_type": "QInt8"
}
}
}