Qwen3.5-2B-SWQ / quantize_config.json
flaviusburca's picture
Initial SWQ release: 5.8357 bpp, gate +0.0100 vs FP8
9449c1b verified
Raw
History Blame Contribute Delete
76.9 kB
{
"quant_method": "surogate_slq",
"format_version": 1,
"quality_track": "primary",
"method": "hybrid_gptq_rtn",
"asymmetric": true,
"group_size": 128,
"bitwidths": [
3,
4,
5,
6,
8
],
"runtime_8bit": "integer",
"layer_meta": {
"model.language_model.layers.0.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0000-71561fa4e0d5-int6.safetensors"
},
"model.language_model.layers.0.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.0.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0001-8d40e2816219-int6.safetensors"
},
"model.language_model.layers.0.mlp.down_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.0.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0002-81a5663d10f9-int4.safetensors"
},
"model.language_model.layers.0.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.0.mlp.gate_proj.weight",
"model.language_model.layers.0.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0003-baee882b8cb9-int6.safetensors"
},
"model.language_model.layers.1.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0004-b9224f2bd0e5-int5.safetensors"
},
"model.language_model.layers.1.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.1.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0005-a010a8529e3e-int6.safetensors"
},
"model.language_model.layers.1.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.1.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0006-4643e16b5baa-int6.safetensors"
},
"model.language_model.layers.1.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.1.mlp.gate_proj.weight",
"model.language_model.layers.1.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0007-42d65d25bc9d-int6.safetensors"
},
"model.language_model.layers.10.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0008-101327d43b53-int6.safetensors"
},
"model.language_model.layers.10.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.10.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0009-66ad9edf9298-int6.safetensors"
},
"model.language_model.layers.10.mlp.down_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.10.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0010-c0137448e42e-int5.safetensors"
},
"model.language_model.layers.10.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.10.mlp.gate_proj.weight",
"model.language_model.layers.10.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0011-5f3697f5a8b0-int5.safetensors"
},
"model.language_model.layers.11.mlp.down_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.11.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0012-04bd5edc9542-int4.safetensors"
},
"model.language_model.layers.11.mlp.gate_up_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.11.mlp.gate_proj.weight",
"model.language_model.layers.11.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0013-c2469e56ca78-int4.safetensors"
},
"model.language_model.layers.11.self_attn.o_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.11.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0014-9138e5fa9728-int6.safetensors"
},
"model.language_model.layers.11.self_attn.qkv_proj": {
"bits": 5,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0015-e9fe59b850de-int5.safetensors"
},
"model.language_model.layers.12.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0016-f9ad788063c5-int5.safetensors"
},
"model.language_model.layers.12.linear_attn.out_proj": {
"bits": 3,
"group_size": 128,
"source_names": [
"model.language_model.layers.12.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0017-7dbfc28e3c9e-int3.safetensors"
},
"model.language_model.layers.12.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.12.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0018-aee158655b4c-int6.safetensors"
},
"model.language_model.layers.12.mlp.gate_up_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.12.mlp.gate_proj.weight",
"model.language_model.layers.12.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0019-b75bf7f9ce51-int4.safetensors"
},
"model.language_model.layers.13.linear_attn.in_proj_qkvz": {
"bits": 4,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0020-9a515716ca81-int4.safetensors"
},
"model.language_model.layers.13.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.13.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0021-b863d689be13-int6.safetensors"
},
"model.language_model.layers.13.mlp.down_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.13.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0022-303f82ff1fbf-int4.safetensors"
},
"model.language_model.layers.13.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.13.mlp.gate_proj.weight",
"model.language_model.layers.13.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0023-ca76416815ef-int5.safetensors"
},
"model.language_model.layers.14.linear_attn.in_proj_qkvz": {
"bits": 4,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0024-f809d137187b-int4.safetensors"
},
"model.language_model.layers.14.linear_attn.out_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.14.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0025-0d13e8c8c32f-int5.safetensors"
},
"model.language_model.layers.14.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.14.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0026-767f3e43f25b-int6.safetensors"
},
"model.language_model.layers.14.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.14.mlp.gate_proj.weight",
"model.language_model.layers.14.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0027-e7466640a2ae-int6.safetensors"
},
"model.language_model.layers.15.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.15.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0028-c1c63aca3301-int6.safetensors"
},
"model.language_model.layers.15.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.15.mlp.gate_proj.weight",
"model.language_model.layers.15.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0029-11c06024045c-int5.safetensors"
},
"model.language_model.layers.15.self_attn.o_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.15.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0030-d2e26ed53669-int5.safetensors"
},
"model.language_model.layers.15.self_attn.qkv_proj": {
"bits": 6,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0031-93051b2df2d2-int6.safetensors"
},
"model.language_model.layers.16.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0032-82d337f7e6bc-int5.safetensors"
},
"model.language_model.layers.16.linear_attn.out_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.16.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0033-0b46aa935ffe-int4.safetensors"
},
"model.language_model.layers.16.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.16.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0034-02fa1f4c31d8-int6.safetensors"
},
"model.language_model.layers.16.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.16.mlp.gate_proj.weight",
"model.language_model.layers.16.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0035-5d64f6f8a9f3-int6.safetensors"
},
"model.language_model.layers.17.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0036-7333467b077c-int5.safetensors"
},
"model.language_model.layers.17.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.17.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0037-39ddb1bd0ddb-int6.safetensors"
},
"model.language_model.layers.17.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.17.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0038-b0f6ed8d7bdb-int6.safetensors"
},
"model.language_model.layers.17.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.17.mlp.gate_proj.weight",
"model.language_model.layers.17.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0039-1cfefb61742a-int6.safetensors"
},
"model.language_model.layers.18.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0040-a176f79c92e5-int5.safetensors"
},
"model.language_model.layers.18.linear_attn.out_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.18.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0041-750a85a23d00-int5.safetensors"
},
"model.language_model.layers.18.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.18.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0042-8078cb9d18fb-int6.safetensors"
},
"model.language_model.layers.18.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.18.mlp.gate_proj.weight",
"model.language_model.layers.18.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0043-228bc6661d49-int6.safetensors"
},
"model.language_model.layers.19.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.19.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0044-a306a6b645f9-int6.safetensors"
},
"model.language_model.layers.19.mlp.gate_up_proj": {
"bits": 3,
"group_size": 128,
"source_names": [
"model.language_model.layers.19.mlp.gate_proj.weight",
"model.language_model.layers.19.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0045-bd05cd830724-int3.safetensors"
},
"model.language_model.layers.19.self_attn.o_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.19.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0046-db9e1452572f-int6.safetensors"
},
"model.language_model.layers.19.self_attn.qkv_proj": {
"bits": 6,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0047-dae2e8b74b5e-int6.safetensors"
},
"model.language_model.layers.2.linear_attn.in_proj_qkvz": {
"bits": 4,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0048-02733a45ba1f-int4.safetensors"
},
"model.language_model.layers.2.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.2.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0049-3b32c5f4eac9-int6.safetensors"
},
"model.language_model.layers.2.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.2.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0050-e33d04c8b946-int6.safetensors"
},
"model.language_model.layers.2.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.2.mlp.gate_proj.weight",
"model.language_model.layers.2.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0051-702da5cd4033-int6.safetensors"
},
"model.language_model.layers.20.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0052-abbe635a6c1d-int6.safetensors"
},
"model.language_model.layers.20.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.20.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0053-9c98b155dee7-int6.safetensors"
},
"model.language_model.layers.20.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.20.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0054-1b6dcac155ae-int6.safetensors"
},
"model.language_model.layers.20.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.20.mlp.gate_proj.weight",
"model.language_model.layers.20.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0055-7f49900a6585-int5.safetensors"
},
"model.language_model.layers.21.linear_attn.in_proj_qkvz": {
"bits": 4,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0056-657c86f3535f-int4.safetensors"
},
"model.language_model.layers.21.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.21.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0057-dccc83c588a1-int6.safetensors"
},
"model.language_model.layers.21.mlp.down_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.21.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0058-04fbdf544dea-int5.safetensors"
},
"model.language_model.layers.21.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.21.mlp.gate_proj.weight",
"model.language_model.layers.21.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0059-3ccc192e6506-int5.safetensors"
},
"model.language_model.layers.22.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0060-060d45b981b2-int6.safetensors"
},
"model.language_model.layers.22.linear_attn.out_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.22.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0061-fe4ea03bcc7d-int5.safetensors"
},
"model.language_model.layers.22.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.22.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0062-821ad23cff4d-int6.safetensors"
},
"model.language_model.layers.22.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.22.mlp.gate_proj.weight",
"model.language_model.layers.22.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0063-061228982b12-int5.safetensors"
},
"model.language_model.layers.23.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.23.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0064-76d02913e3e1-int6.safetensors"
},
"model.language_model.layers.23.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.23.mlp.gate_proj.weight",
"model.language_model.layers.23.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0065-55c52543d9ac-int6.safetensors"
},
"model.language_model.layers.23.self_attn.o_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.23.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0066-b5d0dc4377aa-int6.safetensors"
},
"model.language_model.layers.23.self_attn.qkv_proj": {
"bits": 5,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0067-c11cabd54cc2-int5.safetensors"
},
"model.language_model.layers.3.mlp.down_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.3.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0068-2cfacdd6f18e-int5.safetensors"
},
"model.language_model.layers.3.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.3.mlp.gate_proj.weight",
"model.language_model.layers.3.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0069-a21038d57aeb-int6.safetensors"
},
"model.language_model.layers.3.self_attn.o_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.3.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0070-e66577d9a188-int6.safetensors"
},
"model.language_model.layers.3.self_attn.qkv_proj": {
"bits": 6,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0071-cb52f6a845ed-int6.safetensors"
},
"model.language_model.layers.4.linear_attn.in_proj_qkvz": {
"bits": 4,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0072-ff5debc8e2da-int4.safetensors"
},
"model.language_model.layers.4.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.4.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0073-ed119af61fb4-int6.safetensors"
},
"model.language_model.layers.4.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.4.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0074-6a5e1fa4a01a-int6.safetensors"
},
"model.language_model.layers.4.mlp.gate_up_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.4.mlp.gate_proj.weight",
"model.language_model.layers.4.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0075-e756ce5701aa-int4.safetensors"
},
"model.language_model.layers.5.linear_attn.in_proj_qkvz": {
"bits": 5,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0076-85b270e52168-int5.safetensors"
},
"model.language_model.layers.5.linear_attn.out_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.5.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0077-68f17d9fd665-int5.safetensors"
},
"model.language_model.layers.5.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.5.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0078-245414b84501-int6.safetensors"
},
"model.language_model.layers.5.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.5.mlp.gate_proj.weight",
"model.language_model.layers.5.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0079-bcdda7e27f83-int5.safetensors"
},
"model.language_model.layers.6.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0080-49eb01271051-int6.safetensors"
},
"model.language_model.layers.6.linear_attn.out_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.6.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0081-6e718a97c582-int5.safetensors"
},
"model.language_model.layers.6.mlp.down_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.6.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0082-572bada42f0d-int6.safetensors"
},
"model.language_model.layers.6.mlp.gate_up_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.6.mlp.gate_proj.weight",
"model.language_model.layers.6.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0083-90359b8cbc64-int6.safetensors"
},
"model.language_model.layers.7.mlp.down_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.7.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0084-633af82c7089-int5.safetensors"
},
"model.language_model.layers.7.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.7.mlp.gate_proj.weight",
"model.language_model.layers.7.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0085-a7c090929c33-int5.safetensors"
},
"model.language_model.layers.7.self_attn.o_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.7.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0086-338dc4f792d4-int5.safetensors"
},
"model.language_model.layers.7.self_attn.qkv_proj": {
"bits": 4,
"group_size": 128,
"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": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0087-8a50195a660d-int4.safetensors"
},
"model.language_model.layers.8.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0088-474b6321a087-int6.safetensors"
},
"model.language_model.layers.8.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.8.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0089-e6e8523bbf03-int6.safetensors"
},
"model.language_model.layers.8.mlp.down_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.8.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0090-12ee8d9f0e13-int4.safetensors"
},
"model.language_model.layers.8.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.8.mlp.gate_proj.weight",
"model.language_model.layers.8.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0091-be7c1adaccf4-int5.safetensors"
},
"model.language_model.layers.9.linear_attn.in_proj_qkvz": {
"bits": 6,
"group_size": 128,
"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,
2048,
2048
],
"in_features": 2048,
"out_features": 8192,
"candidate": "unit-0092-78de9aea31ca-int6.safetensors"
},
"model.language_model.layers.9.linear_attn.out_proj": {
"bits": 6,
"group_size": 128,
"source_names": [
"model.language_model.layers.9.linear_attn.out_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0093-d7f3f7a0a618-int6.safetensors"
},
"model.language_model.layers.9.mlp.down_proj": {
"bits": 4,
"group_size": 128,
"source_names": [
"model.language_model.layers.9.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0094-c62095daa0ad-int4.safetensors"
},
"model.language_model.layers.9.mlp.gate_up_proj": {
"bits": 5,
"group_size": 128,
"source_names": [
"model.language_model.layers.9.mlp.gate_proj.weight",
"model.language_model.layers.9.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0095-cff26e1403df-int5.safetensors"
},
"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-0096-1509ccecde6a-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-0097-12841d14a3b9-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-0098-678f84241377-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-0099-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-0100-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-0101-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-0102-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-0103-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-0104-7264499e4947-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-0105-66c2bc0dc07b-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-0106-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-0107-1c60ed11b7a0-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-0108-1ddf25873145-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-0109-9c893daff9e9-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-0110-941ec1ddb810-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-0111-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-0112-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-0113-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-0114-48547f74f6cc-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-0115-4de972bb4f1b-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-0116-349e5622965f-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-0117-05660b332cbc-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-0118-2f2b68c52101-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-0119-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-0120-2c3afacb044a-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-0121-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-0122-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-0123-203d94db1bf9-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-0124-78db9f9e6d1e-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-0125-176a99fca0b3-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-0126-0d59f62e72e8-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-0127-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-0128-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-0129-38faada5043c-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-0130-4775e120874f-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-0131-968964931025-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-0132-5d6743f37757-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-0133-37bb9e81b701-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-0134-9724aa91564d-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-0135-1cbe44debe73-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-0136-1dede7f7e096-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-0137-64c52ea6cbe3-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-0138-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-0139-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-0140-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-0141-530359111129-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-0142-2f5734ea658e-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-0143-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-0144-32de5ff04751-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-0145-80b03bd45f33-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-0146-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-0147-26816b847b50-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-0148-5268bc8acb20-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-0149-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-0150-20f2142ac222-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-0151-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-0152-5b8df0449351-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-0153-15a6a71c7caf-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-0154-813c2f94dbce-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-0155-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-0156-24338fcfa0a0-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-0157-77bb2f91f182-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-0158-2f87bf92ff4d-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-0159-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-0160-693d47f0a0aa-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-0161-5d031fa92038-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-0162-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-0163-438cf5e2595e-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-0164-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-0165-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-0166-56d3adda7f18-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-0167-5cd51a2921c6-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-0168-69760cf67c53-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-0169-68143911722c-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-0170-338e2facf804-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-0171-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-0172-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-0173-8bf53b2c472b-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-0174-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-0175-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-0176-60f7be1b5d5d-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-0177-73f240ed8e47-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-0178-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-0179-7d70ad25adfd-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-0180-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-0181-574a8ccf67a0-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-0182-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-0183-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-0184-9dee46109ef6-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-0185-28414763f199-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-0186-0c591f99cc1d-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-0187-5382d153fb25-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-0188-8711cb24f2c2-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-0189-9b7bd3d15eb3-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-0190-88f41671f902-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-0191-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-0192-3af477e41f35-int8.safetensors"
},
"model.visual.merger.linear_fc2": {
"bits": 8,
"group_size": 128,
"source_names": [
"model.visual.merger.linear_fc2.weight"
],
"output_splits": [
2048
],
"in_features": 4096,
"out_features": 2048,
"candidate": "unit-0193-0881cfb9d592-int8.safetensors"
},
"mtp.fc": {
"bits": 8,
"group_size": 128,
"source_names": [
"mtp.fc.weight"
],
"output_splits": [
2048
],
"in_features": 4096,
"out_features": 2048,
"candidate": "unit-0194-7e6f34ca4b04-int8.safetensors"
},
"mtp.layers.0.mlp.down_proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"mtp.layers.0.mlp.down_proj.weight"
],
"output_splits": [
2048
],
"in_features": 6144,
"out_features": 2048,
"candidate": "unit-0195-8645579d9447-int8.safetensors"
},
"mtp.layers.0.mlp.gate_up_proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"mtp.layers.0.mlp.gate_proj.weight",
"mtp.layers.0.mlp.up_proj.weight"
],
"output_splits": [
6144,
6144
],
"in_features": 2048,
"out_features": 12288,
"candidate": "unit-0196-cbb761be6bb4-int8.safetensors"
},
"mtp.layers.0.self_attn.o_proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"mtp.layers.0.self_attn.o_proj.weight"
],
"output_splits": [
2048
],
"in_features": 2048,
"out_features": 2048,
"candidate": "unit-0197-810ea1ccf166-int8.safetensors"
},
"mtp.layers.0.self_attn.qkv_proj": {
"bits": 8,
"group_size": 128,
"source_names": [
"mtp.layers.0.self_attn.q_proj.weight",
"mtp.layers.0.self_attn.k_proj.weight",
"mtp.layers.0.self_attn.v_proj.weight"
],
"output_splits": [
4096,
512,
512
],
"in_features": 2048,
"out_features": 5120,
"candidate": "unit-0198-d0295a375904-int8.safetensors"
}
},
"bf16_projections": [],
"deployment": {
"tensor_payload_bytes": 1676739736,
"whole_model_bpp": 5.898639411346412,
"rate_denominator_parameters": 2274069824,
"stripped_prefixes": []
},
"embedding": {
"kind": "affine_embedding",
"prefix": "model.language_model.embed_tokens",
"source_name": "model.language_model.embed_tokens.weight",
"bits": 5,
"group_size": 128,
"num_embeddings": 248320,
"embedding_dim": 2048,
"packed_columns": 1280,
"payload_bytes": 333742100,
"source_payload_bytes": 1017118720,
"tied_lm_head": true,
"rate_status": "fixed"
}
}