changwangss's picture
Add full-encoder MXFP8 RTN FLUX.2 model and tested inference examples
6ec783c verified
Raw History Blame Contribute Delete
621 Bytes
{
"quant_method": "auto-round",
"packing_format": "auto_round:llm_compressor",
"bits": 8,
"group_size": 32,
"sym": true,
"data_type": "mx_fp",
"iters": 0,
"model_free": true,
"autoround_version": "0.17.0",
"enable_quanted_input": false,
"block_name_to_quantize": "language_model.model.layers,vision_tower.transformer.layers",
"act_bits": 8,
"act_data_type": "mx_fp",
"act_dynamic": true,
"act_group_size": 32,
"act_sym": true,
"extra_config": {
"language_model.lm_head": {
"bits": 16,
"data_type": "float",
"act_bits": 16,
"act_data_type": "float"
}
}
}