Files
octen-embedding-8b-w4a16/quantization_config.json

15 lines
314 B
JSON
Raw Permalink Normal View History

{
"bits": 4,
"data_type": "int",
"group_size": 128,
"sym": true,
"amp": false,
"batch_size": 1,
"gradient_accumulate_steps": 8,
"low_gpu_mem_usage": true,
"nsamples": 8,
"seqlen": 512,
"autoround_version": "0.12.2",
"quant_method": "auto-round",
"packing_format": "auto_round:auto_gptq"
}