kev-9b-W4A16 / quantization_config.json
bowmanslayer's picture
Add quantization_config.json
3a4297b verified
Raw History Blame Contribute Delete
3.4 kB
{
"bits": 4,
"data_type": "int",
"group_size": 128,
"sym": true,
"batch_size": 4,
"low_gpu_mem_usage": true,
"nsamples": 256,
"seqlen": 512,
"autoround_version": "0.12.3",
"quant_method": "auto-round",
"packing_format": "auto_round:auto_gptq",
"extra_config": {
"layers.0.linear_attn.in_proj_b": {
"bits": 16
},
"layers.0.linear_attn.in_proj_a": {
"bits": 16
},
"layers.1.linear_attn.in_proj_b": {
"bits": 16
},
"layers.1.linear_attn.in_proj_a": {
"bits": 16
},
"layers.2.linear_attn.in_proj_b": {
"bits": 16
},
"layers.2.linear_attn.in_proj_a": {
"bits": 16
},
"layers.4.linear_attn.in_proj_b": {
"bits": 16
},
"layers.4.linear_attn.in_proj_a": {
"bits": 16
},
"layers.5.linear_attn.in_proj_b": {
"bits": 16
},
"layers.5.linear_attn.in_proj_a": {
"bits": 16
},
"layers.6.linear_attn.in_proj_b": {
"bits": 16
},
"layers.6.linear_attn.in_proj_a": {
"bits": 16
},
"layers.8.linear_attn.in_proj_b": {
"bits": 16
},
"layers.8.linear_attn.in_proj_a": {
"bits": 16
},
"layers.9.linear_attn.in_proj_b": {
"bits": 16
},
"layers.9.linear_attn.in_proj_a": {
"bits": 16
},
"layers.10.linear_attn.in_proj_b": {
"bits": 16
},
"layers.10.linear_attn.in_proj_a": {
"bits": 16
},
"layers.12.linear_attn.in_proj_b": {
"bits": 16
},
"layers.12.linear_attn.in_proj_a": {
"bits": 16
},
"layers.13.linear_attn.in_proj_b": {
"bits": 16
},
"layers.13.linear_attn.in_proj_a": {
"bits": 16
},
"layers.14.linear_attn.in_proj_b": {
"bits": 16
},
"layers.14.linear_attn.in_proj_a": {
"bits": 16
},
"layers.16.linear_attn.in_proj_b": {
"bits": 16
},
"layers.16.linear_attn.in_proj_a": {
"bits": 16
},
"layers.17.linear_attn.in_proj_b": {
"bits": 16
},
"layers.17.linear_attn.in_proj_a": {
"bits": 16
},
"layers.18.linear_attn.in_proj_b": {
"bits": 16
},
"layers.18.linear_attn.in_proj_a": {
"bits": 16
},
"layers.20.linear_attn.in_proj_b": {
"bits": 16
},
"layers.20.linear_attn.in_proj_a": {
"bits": 16
},
"layers.21.linear_attn.in_proj_b": {
"bits": 16
},
"layers.21.linear_attn.in_proj_a": {
"bits": 16
},
"layers.22.linear_attn.in_proj_b": {
"bits": 16
},
"layers.22.linear_attn.in_proj_a": {
"bits": 16
},
"layers.24.linear_attn.in_proj_b": {
"bits": 16
},
"layers.24.linear_attn.in_proj_a": {
"bits": 16
},
"layers.25.linear_attn.in_proj_b": {
"bits": 16
},
"layers.25.linear_attn.in_proj_a": {
"bits": 16
},
"layers.26.linear_attn.in_proj_b": {
"bits": 16
},
"layers.26.linear_attn.in_proj_a": {
"bits": 16
},
"layers.28.linear_attn.in_proj_b": {
"bits": 16
},
"layers.28.linear_attn.in_proj_a": {
"bits": 16
},
"layers.29.linear_attn.in_proj_b": {
"bits": 16
},
"layers.29.linear_attn.in_proj_a": {
"bits": 16
},
"layers.30.linear_attn.in_proj_b": {
"bits": 16
},
"layers.30.linear_attn.in_proj_a": {
"bits": 16
}
}
}