Qwen3.5-4B-SWQ / quantize_config.json
flaviusburca's picture
SWQ 5.66 bpp multimodal: 3.3 GB, no benchmark separates it from BF16
a17d385 verified
Raw
History Blame Contribute Delete
84.5 kB
{
"quant_method": "surogate_slq",
"format_version": 1,
"quality_track": "primary",
"method": "hierarchical_gptq_refit",
"asymmetric": true,
"group_size": 128,
"bitwidths": [
4,
5,
6,
8
],
"bf16_projections": [],
"deployment": {
"tensor_payload_bytes": 2977349136,
"whole_model_bpp": 5.663386018724775,
"rate_denominator_parameters": 4205751296,
"vision_payload_bytes": 345495224,
"vision_rate_status": "excluded",
"stripped_prefixes": []
},
"layer_meta": {
"model.language_model.layers.0.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.0.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.0.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.0.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.0.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.0.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.0.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.0.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.0.mlp.gate_proj.weight",
"model.language_model.layers.0.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.1.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.1.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.1.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.1.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.1.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.1.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.1.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.1.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.1.mlp.gate_proj.weight",
"model.language_model.layers.1.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.10.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.10.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.10.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.10.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.10.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.10.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.10.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.10.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.10.mlp.gate_proj.weight",
"model.language_model.layers.10.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.11.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.11.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.11.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.11.mlp.gate_proj.weight",
"model.language_model.layers.11.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.11.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.11.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.11.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.11.self_attn.q_proj.weight",
"model.language_model.layers.11.self_attn.k_proj.weight",
"model.language_model.layers.11.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.12.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.12.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.12.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.12.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.12.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.12.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.12.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.12.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.12.mlp.gate_proj.weight",
"model.language_model.layers.12.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.13.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.13.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.13.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.13.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.13.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.13.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.13.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.13.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.13.mlp.gate_proj.weight",
"model.language_model.layers.13.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.14.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.14.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.14.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.14.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.14.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.14.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.14.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.14.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.14.mlp.gate_proj.weight",
"model.language_model.layers.14.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.15.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.15.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.15.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.15.mlp.gate_proj.weight",
"model.language_model.layers.15.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.15.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.15.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.15.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.15.self_attn.q_proj.weight",
"model.language_model.layers.15.self_attn.k_proj.weight",
"model.language_model.layers.15.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.16.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.16.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.16.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.16.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.16.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.16.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.16.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.16.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.16.mlp.gate_proj.weight",
"model.language_model.layers.16.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.17.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.17.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.17.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.17.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.17.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.17.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.17.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.17.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.17.mlp.gate_proj.weight",
"model.language_model.layers.17.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.18.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.18.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.18.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.18.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.18.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.18.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.18.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.18.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.18.mlp.gate_proj.weight",
"model.language_model.layers.18.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.19.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.19.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.19.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.19.mlp.gate_proj.weight",
"model.language_model.layers.19.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.19.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.19.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.19.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.19.self_attn.q_proj.weight",
"model.language_model.layers.19.self_attn.k_proj.weight",
"model.language_model.layers.19.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.2.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.2.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.2.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.2.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.2.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.2.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.2.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.2.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.2.mlp.gate_proj.weight",
"model.language_model.layers.2.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.20.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.20.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.20.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.20.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.20.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.20.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.20.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.20.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.20.mlp.gate_proj.weight",
"model.language_model.layers.20.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.21.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.21.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.21.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.21.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.21.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.21.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.21.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.21.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.21.mlp.gate_proj.weight",
"model.language_model.layers.21.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.22.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.22.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.22.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.22.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.22.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.22.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.22.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.22.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.22.mlp.gate_proj.weight",
"model.language_model.layers.22.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.23.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.23.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.23.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.23.mlp.gate_proj.weight",
"model.language_model.layers.23.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.23.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.23.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.23.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.23.self_attn.q_proj.weight",
"model.language_model.layers.23.self_attn.k_proj.weight",
"model.language_model.layers.23.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.24.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.24.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.24.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.24.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.24.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.24.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.24.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.24.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.24.mlp.gate_proj.weight",
"model.language_model.layers.24.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.25.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.25.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.25.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.25.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.25.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.25.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.25.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.25.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.25.mlp.gate_proj.weight",
"model.language_model.layers.25.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.26.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.26.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.26.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.26.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.26.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.26.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.26.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.26.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.26.mlp.gate_proj.weight",
"model.language_model.layers.26.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.27.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.27.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.27.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.27.mlp.gate_proj.weight",
"model.language_model.layers.27.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.27.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.27.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.27.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.27.self_attn.q_proj.weight",
"model.language_model.layers.27.self_attn.k_proj.weight",
"model.language_model.layers.27.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.28.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.28.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.28.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.28.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.28.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.28.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.28.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.28.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.28.mlp.gate_proj.weight",
"model.language_model.layers.28.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.29.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.29.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.29.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.29.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.29.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.29.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.29.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.29.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.29.mlp.gate_proj.weight",
"model.language_model.layers.29.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.3.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.3.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.3.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.3.mlp.gate_proj.weight",
"model.language_model.layers.3.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.3.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.3.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.3.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.3.self_attn.q_proj.weight",
"model.language_model.layers.3.self_attn.k_proj.weight",
"model.language_model.layers.3.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.30.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.30.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.30.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.30.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.30.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.30.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.30.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.30.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.30.mlp.gate_proj.weight",
"model.language_model.layers.30.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.31.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.31.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.31.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.31.mlp.gate_proj.weight",
"model.language_model.layers.31.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.31.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.31.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.31.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.31.self_attn.q_proj.weight",
"model.language_model.layers.31.self_attn.k_proj.weight",
"model.language_model.layers.31.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.4.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.4.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.4.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.4.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.4.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.4.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.4.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.4.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.4.mlp.gate_proj.weight",
"model.language_model.layers.4.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.5.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.5.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.5.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.5.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.5.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.5.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.5.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.5.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.5.mlp.gate_proj.weight",
"model.language_model.layers.5.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.6.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.6.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.6.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.6.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.6.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.6.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.6.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.6.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.6.mlp.gate_proj.weight",
"model.language_model.layers.6.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.7.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.7.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.7.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.7.mlp.gate_proj.weight",
"model.language_model.layers.7.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.7.self_attn.o_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.7.self_attn.o_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.7.self_attn.qkv_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.7.self_attn.q_proj.weight",
"model.language_model.layers.7.self_attn.k_proj.weight",
"model.language_model.layers.7.self_attn.v_proj.weight"
],
"output_splits": [
8192,
1024,
1024
],
"in_features": 2560,
"out_features": 10240
},
"model.language_model.layers.8.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.8.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.8.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.8.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.8.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.8.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.8.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.8.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.8.mlp.gate_proj.weight",
"model.language_model.layers.8.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.language_model.layers.9.linear_attn.in_proj_qkvz": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.9.linear_attn.in_proj_qkv.weight",
"model.language_model.layers.9.linear_attn.in_proj_z.weight"
],
"output_splits": [
2048,
2048,
4096,
4096
],
"in_features": 2560,
"out_features": 12288
},
"model.language_model.layers.9.linear_attn.out_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.9.linear_attn.out_proj.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560
},
"model.language_model.layers.9.mlp.down_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.9.mlp.down_proj.weight"
],
"output_splits": [
2560
],
"in_features": 9216,
"out_features": 2560
},
"model.language_model.layers.9.mlp.gate_up_proj": {
"bits": 5,
"runtime_format": "hier",
"group_size": 32,
"source_names": [
"model.language_model.layers.9.mlp.gate_proj.weight",
"model.language_model.layers.9.mlp.up_proj.weight"
],
"output_splits": [
9216,
9216
],
"in_features": 2560,
"out_features": 18432
},
"model.visual.blocks.0.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.0.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0128-1509ccecde6a-int8.safetensors"
},
"model.visual.blocks.0.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.0.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0129-12841d14a3b9-int8.safetensors"
},
"model.visual.blocks.0.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.0.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0130-678f84241377-int8.safetensors"
},
"model.visual.blocks.0.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.0.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0131-d369cf31dd8a-int8.safetensors"
},
"model.visual.blocks.1.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.1.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0132-f08b997bbd8e-int8.safetensors"
},
"model.visual.blocks.1.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.1.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0133-cbba755d8466-int8.safetensors"
},
"model.visual.blocks.1.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.1.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0134-d5ce36f05818-int8.safetensors"
},
"model.visual.blocks.1.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.1.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0135-eee2104df7d3-int8.safetensors"
},
"model.visual.blocks.10.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.10.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0136-7264499e4947-int8.safetensors"
},
"model.visual.blocks.10.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.10.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0137-66c2bc0dc07b-int8.safetensors"
},
"model.visual.blocks.10.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.10.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0138-c013ba7ac38b-int8.safetensors"
},
"model.visual.blocks.10.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.10.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0139-1c60ed11b7a0-int8.safetensors"
},
"model.visual.blocks.11.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.11.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0140-1ddf25873145-int8.safetensors"
},
"model.visual.blocks.11.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.11.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0141-9c893daff9e9-int8.safetensors"
},
"model.visual.blocks.11.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.11.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0142-941ec1ddb810-int8.safetensors"
},
"model.visual.blocks.11.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.11.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0143-f36c92fb723a-int8.safetensors"
},
"model.visual.blocks.12.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.12.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0144-b3816ff05d08-int8.safetensors"
},
"model.visual.blocks.12.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.12.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0145-ec50e3c4b370-int8.safetensors"
},
"model.visual.blocks.12.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.12.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0146-48547f74f6cc-int8.safetensors"
},
"model.visual.blocks.12.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.12.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0147-4de972bb4f1b-int8.safetensors"
},
"model.visual.blocks.13.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.13.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0148-349e5622965f-int8.safetensors"
},
"model.visual.blocks.13.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.13.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0149-05660b332cbc-int8.safetensors"
},
"model.visual.blocks.13.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.13.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0150-2f2b68c52101-int8.safetensors"
},
"model.visual.blocks.13.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.13.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0151-abd33a4c0c66-int8.safetensors"
},
"model.visual.blocks.14.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.14.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0152-2c3afacb044a-int8.safetensors"
},
"model.visual.blocks.14.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.14.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0153-f253f8de9ab8-int8.safetensors"
},
"model.visual.blocks.14.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.14.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0154-b881011b6eae-int8.safetensors"
},
"model.visual.blocks.14.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.14.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0155-203d94db1bf9-int8.safetensors"
},
"model.visual.blocks.15.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.15.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0156-78db9f9e6d1e-int8.safetensors"
},
"model.visual.blocks.15.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.15.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0157-176a99fca0b3-int8.safetensors"
},
"model.visual.blocks.15.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.15.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0158-0d59f62e72e8-int8.safetensors"
},
"model.visual.blocks.15.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.15.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0159-deec42f17db0-int8.safetensors"
},
"model.visual.blocks.16.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.16.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0160-fbd5c9096d76-int8.safetensors"
},
"model.visual.blocks.16.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.16.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0161-38faada5043c-int8.safetensors"
},
"model.visual.blocks.16.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.16.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0162-4775e120874f-int8.safetensors"
},
"model.visual.blocks.16.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.16.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0163-968964931025-int8.safetensors"
},
"model.visual.blocks.17.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.17.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0164-5d6743f37757-int8.safetensors"
},
"model.visual.blocks.17.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.17.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0165-37bb9e81b701-int8.safetensors"
},
"model.visual.blocks.17.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.17.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0166-9724aa91564d-int8.safetensors"
},
"model.visual.blocks.17.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.17.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0167-1cbe44debe73-int8.safetensors"
},
"model.visual.blocks.18.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.18.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0168-1dede7f7e096-int8.safetensors"
},
"model.visual.blocks.18.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.18.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0169-64c52ea6cbe3-int8.safetensors"
},
"model.visual.blocks.18.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.18.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0170-f7c0a6ab6f23-int8.safetensors"
},
"model.visual.blocks.18.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.18.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0171-fee90f442f70-int8.safetensors"
},
"model.visual.blocks.19.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.19.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0172-cab6876bc162-int8.safetensors"
},
"model.visual.blocks.19.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.19.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0173-530359111129-int8.safetensors"
},
"model.visual.blocks.19.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.19.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0174-2f5734ea658e-int8.safetensors"
},
"model.visual.blocks.19.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.19.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0175-d0b9d917d2aa-int8.safetensors"
},
"model.visual.blocks.2.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.2.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0176-32de5ff04751-int8.safetensors"
},
"model.visual.blocks.2.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.2.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0177-80b03bd45f33-int8.safetensors"
},
"model.visual.blocks.2.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.2.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0178-d84c99b6e16d-int8.safetensors"
},
"model.visual.blocks.2.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.2.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0179-26816b847b50-int8.safetensors"
},
"model.visual.blocks.20.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.20.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0180-5268bc8acb20-int8.safetensors"
},
"model.visual.blocks.20.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.20.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0181-d1e0cc36a78c-int8.safetensors"
},
"model.visual.blocks.20.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.20.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0182-20f2142ac222-int8.safetensors"
},
"model.visual.blocks.20.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.20.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0183-c6e2b6c62725-int8.safetensors"
},
"model.visual.blocks.21.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.21.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0184-5b8df0449351-int8.safetensors"
},
"model.visual.blocks.21.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.21.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0185-15a6a71c7caf-int8.safetensors"
},
"model.visual.blocks.21.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.21.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0186-813c2f94dbce-int8.safetensors"
},
"model.visual.blocks.21.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.21.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0187-ea54e1eb6e1b-int8.safetensors"
},
"model.visual.blocks.22.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.22.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0188-24338fcfa0a0-int8.safetensors"
},
"model.visual.blocks.22.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.22.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0189-77bb2f91f182-int8.safetensors"
},
"model.visual.blocks.22.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.22.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0190-2f87bf92ff4d-int8.safetensors"
},
"model.visual.blocks.22.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.22.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0191-a8873a024775-int8.safetensors"
},
"model.visual.blocks.23.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.23.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0192-693d47f0a0aa-int8.safetensors"
},
"model.visual.blocks.23.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.23.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0193-5d031fa92038-int8.safetensors"
},
"model.visual.blocks.23.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.23.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0194-a403d09d1b2a-int8.safetensors"
},
"model.visual.blocks.23.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.23.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0195-438cf5e2595e-int8.safetensors"
},
"model.visual.blocks.3.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.3.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0196-a405739514ba-int8.safetensors"
},
"model.visual.blocks.3.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.3.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0197-f1ee64957ad4-int8.safetensors"
},
"model.visual.blocks.3.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.3.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0198-56d3adda7f18-int8.safetensors"
},
"model.visual.blocks.3.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.3.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0199-5cd51a2921c6-int8.safetensors"
},
"model.visual.blocks.4.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.4.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0200-69760cf67c53-int8.safetensors"
},
"model.visual.blocks.4.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.4.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0201-68143911722c-int8.safetensors"
},
"model.visual.blocks.4.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.4.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0202-338e2facf804-int8.safetensors"
},
"model.visual.blocks.4.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.4.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0203-f7ffede5ba42-int8.safetensors"
},
"model.visual.blocks.5.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.5.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0204-a916d58cd8e2-int8.safetensors"
},
"model.visual.blocks.5.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.5.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0205-8bf53b2c472b-int8.safetensors"
},
"model.visual.blocks.5.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.5.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0206-c02581ffc039-int8.safetensors"
},
"model.visual.blocks.5.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.5.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0207-a2700be648ca-int8.safetensors"
},
"model.visual.blocks.6.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.6.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0208-60f7be1b5d5d-int8.safetensors"
},
"model.visual.blocks.6.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.6.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0209-73f240ed8e47-int8.safetensors"
},
"model.visual.blocks.6.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.6.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0210-f64ea3a2685c-int8.safetensors"
},
"model.visual.blocks.6.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.6.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0211-7d70ad25adfd-int8.safetensors"
},
"model.visual.blocks.7.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.7.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0212-f6cde3759a14-int8.safetensors"
},
"model.visual.blocks.7.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.7.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0213-574a8ccf67a0-int8.safetensors"
},
"model.visual.blocks.7.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.7.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0214-ca5e8bc7003f-int8.safetensors"
},
"model.visual.blocks.7.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.7.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0215-ad5199e77ecc-int8.safetensors"
},
"model.visual.blocks.8.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.8.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0216-9dee46109ef6-int8.safetensors"
},
"model.visual.blocks.8.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.8.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0217-28414763f199-int8.safetensors"
},
"model.visual.blocks.8.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.8.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0218-0c591f99cc1d-int8.safetensors"
},
"model.visual.blocks.8.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.8.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0219-5382d153fb25-int8.safetensors"
},
"model.visual.blocks.9.attn.proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.9.attn.proj.weight"
],
"output_splits": [
1024
],
"in_features": 1024,
"out_features": 1024,
"candidate": "unit-0220-8711cb24f2c2-int8.safetensors"
},
"model.visual.blocks.9.attn.qkv": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.9.attn.qkv.weight"
],
"output_splits": [
1024,
1024,
1024
],
"in_features": 1024,
"out_features": 3072,
"candidate": "unit-0221-9b7bd3d15eb3-int8.safetensors"
},
"model.visual.blocks.9.mlp.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.9.mlp.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 1024,
"out_features": 4096,
"candidate": "unit-0222-88f41671f902-int8.safetensors"
},
"model.visual.blocks.9.mlp.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.blocks.9.mlp.linear_fc2.weight"
],
"output_splits": [
1024
],
"in_features": 4096,
"out_features": 1024,
"candidate": "unit-0223-f23948988fad-int8.safetensors"
},
"model.visual.merger.linear_fc1": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.merger.linear_fc1.weight"
],
"output_splits": [
4096
],
"in_features": 4096,
"out_features": 4096,
"candidate": "unit-0224-3af477e41f35-int8.safetensors"
},
"model.visual.merger.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.merger.linear_fc2.weight"
],
"output_splits": [
2560
],
"in_features": 4096,
"out_features": 2560,
"candidate": "unit-0225-0881cfb9d592-int8.safetensors"
}
},
"embedding": {
"kind": "hier_embedding",
"prefix": "model.language_model.embed_tokens",
"source_name": "model.language_model.embed_tokens.weight",
"bits": 6,
"num_embeddings": 248320,
"embedding_dim": 2560,
"payload_bytes": 516505600,
"tied_lm_head": true,
"rate_status": "fixed"
}
}