Inkling-975B-Alis-MLX-Dynamic-3.7bpw / conversion_plan.json
avlp12's picture
Upload conversion_plan.json with huggingface_hub
5d67e06 verified
Raw
History Blame Contribute Delete
487 kB
{
"manifest": {
"model.audio.encoder.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 15728640,
"rule": "preserve",
"shape": [
1280,
6144
],
"shard": "model-00091-of-00108.safetensors",
"source_bytes": 15728640
},
"model.audio.final_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.embed.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 2470182912,
"rule": "preserve",
"shape": [
201024,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 2470182912
},
"model.llm.embed_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.0.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.0.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00030-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.0.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.0.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.0.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.0.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.0.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.0.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.0.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.0.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.0.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.0.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.0.mlp.global_scale": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 2,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 2
},
"model.llm.layers.0.mlp.w13_dn.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 603979776,
"rule": "preserve",
"shape": [
49152,
6144
],
"shard": "model-00055-of-00108.safetensors",
"source_bytes": 603979776
},
"model.llm.layers.0.mlp.w2_md.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 301989888,
"rule": "preserve",
"shape": [
6144,
24576
],
"shard": "model-00029-of-00108.safetensors",
"source_bytes": 301989888
},
"model.llm.layers.0.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.0.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.1.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.1.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.1.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.1.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.1.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.1.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.1.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.1.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.1.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.1.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.1.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.1.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.1.mlp.global_scale": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 2,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 2
},
"model.llm.layers.1.mlp.w13_dn.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 603979776,
"rule": "preserve",
"shape": [
49152,
6144
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 603979776
},
"model.llm.layers.1.mlp.w2_md.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 301989888,
"rule": "preserve",
"shape": [
6144,
24576
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 301989888
},
"model.llm.layers.1.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.1.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.10.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.10.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.10.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.10.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.10.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00055-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.10.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.10.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.10.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.10.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.10.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.10.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.10.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.10.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00060-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.10.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.10.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.10.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.10.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.10.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.10.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.10.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00066-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.10.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.11.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.11.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.11.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.11.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.11.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.11.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.11.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.11.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.11.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.11.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.11.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.11.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.11.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00014-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.11.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.11.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.11.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.11.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.11.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.11.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.11.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.11.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.12.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.12.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.12.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.12.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.12.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.12.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.12.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.12.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.12.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.12.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.12.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.12.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.12.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00063-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.12.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.12.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.12.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.12.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.12.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.12.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00005-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.12.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.12.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.13.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.13.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.13.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.13.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.13.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.13.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.13.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.13.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.13.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.13.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.13.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.13.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.13.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00066-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.13.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.13.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00038-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.13.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.13.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.13.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.13.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.13.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.13.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.14.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.14.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.14.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.14.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.14.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.14.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.14.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.14.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.14.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.14.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.14.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.14.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.14.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00095-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.14.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00017-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.14.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.14.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.14.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.14.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.14.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.14.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.14.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.15.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.15.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.15.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00103-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.15.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.15.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.15.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.15.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.15.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.15.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.15.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.15.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.15.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.15.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00015-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.15.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.15.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.15.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.15.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.15.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.15.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.15.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.15.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.16.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.16.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.16.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.16.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.16.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.16.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.16.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.16.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.16.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.16.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.16.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00091-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.16.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.16.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00051-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.16.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.16.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.16.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.16.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.16.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.16.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.16.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.16.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.17.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.17.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.17.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.17.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.17.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.17.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00017-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.17.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.17.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.17.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00054-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.17.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.17.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.17.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.17.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00035-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.17.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.17.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.17.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.17.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00015-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.17.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.17.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.17.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.17.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.18.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.18.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.18.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.18.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.18.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.18.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.18.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00095-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.18.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.18.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.18.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.18.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.18.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.18.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.18.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.18.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.18.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.18.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.18.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.18.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.18.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.18.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.19.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.19.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.19.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.19.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00086-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.19.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.19.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.19.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.19.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.19.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.19.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.19.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.19.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.19.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.19.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.19.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.19.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.19.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.19.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.19.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.19.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.19.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.2.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.2.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.2.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.2.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.2.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.2.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.2.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.2.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00065-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.2.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.2.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.2.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.2.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.2.mlp.experts.w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 19327352832,
"rule": "preserve",
"shape": [
256,
6144,
6144
],
"shard": "model-00107-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.2.mlp.experts.w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 9663676416,
"rule": "preserve",
"shape": [
256,
6144,
3072
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.2.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.2.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.2.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.2.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.2.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.2.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.2.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.20.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.20.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.20.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00103-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.20.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.20.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.20.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.20.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.20.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.20.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.20.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.20.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.20.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.20.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.20.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.20.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.20.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.20.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.20.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.20.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.20.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.20.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.21.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.21.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.21.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.21.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.21.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.21.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.21.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.21.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.21.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.21.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.21.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.21.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.21.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.21.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00047-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.21.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.21.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.21.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.21.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.21.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.21.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.21.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.22.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.22.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.22.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.22.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.22.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.22.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.22.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.22.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.22.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.22.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.22.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.22.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.22.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.22.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.22.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.22.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.22.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.22.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.22.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.22.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.22.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.23.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.23.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.23.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.23.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.23.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.23.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.23.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.23.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.23.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.23.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.23.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.23.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.23.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00065-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.23.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.23.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.23.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.23.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.23.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.23.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.23.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.23.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.24.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.24.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00075-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.24.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.24.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.24.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00091-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.24.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.24.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.24.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.24.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.24.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.24.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.24.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.24.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.24.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.24.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.24.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.24.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.24.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.24.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.24.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.24.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.25.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.25.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.25.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.25.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.25.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.25.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.25.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.25.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.25.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.25.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.25.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.25.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.25.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.25.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.25.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.25.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.25.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.25.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.25.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.25.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.25.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.26.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.26.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.26.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.26.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.26.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.26.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.26.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.26.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.26.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.26.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.26.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.26.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.26.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.26.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.26.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00068-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.26.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.26.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.26.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.26.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.26.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.26.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.27.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.27.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.27.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.27.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.27.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.27.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.27.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.27.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.27.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.27.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.27.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.27.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.27.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.27.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.27.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.27.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.27.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.27.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.27.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.27.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.27.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.28.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.28.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.28.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.28.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.28.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.28.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.28.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.28.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.28.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.28.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.28.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.28.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.28.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.28.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00068-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.28.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.28.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.28.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.28.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.28.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.28.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.28.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.29.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.29.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.29.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00055-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.29.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.29.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.29.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.29.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.29.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.29.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.29.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.29.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00017-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.29.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.29.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.29.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.29.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.29.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.29.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00066-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.29.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.29.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.29.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.29.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.3.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.3.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.3.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.3.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.3.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.3.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.3.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.3.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.3.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.3.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.3.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.3.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.3.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.3.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.3.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.3.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.3.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.3.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.3.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.3.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.3.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.30.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.30.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.30.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.30.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.30.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.30.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.30.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.30.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.30.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.30.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.30.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.30.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.30.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.30.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00077-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.30.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.30.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.30.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.30.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.30.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.30.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.30.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.31.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.31.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.31.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.31.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.31.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.31.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.31.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.31.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.31.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.31.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00060-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.31.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.31.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.31.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.31.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.31.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.31.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.31.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.31.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.31.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.31.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.31.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.32.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.32.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.32.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.32.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.32.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.32.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.32.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.32.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00095-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.32.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.32.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.32.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.32.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.32.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.32.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.32.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.32.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.32.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.32.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00091-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.32.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.32.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.32.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00095-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.33.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.33.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.33.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.33.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.33.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.33.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.33.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.33.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.33.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.33.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.33.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.33.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.33.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00054-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.33.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.33.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.33.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.33.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.33.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.33.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.33.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.33.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.34.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.34.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.34.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.34.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.34.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.34.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.34.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.34.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.34.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.34.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.34.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.34.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.34.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00038-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.34.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.34.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.34.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.34.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.34.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.34.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.34.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.34.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.35.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.35.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.35.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.35.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.35.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.35.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00035-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.35.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.35.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.35.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.35.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.35.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.35.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.35.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00097-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.35.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.35.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.35.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.35.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.35.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.35.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.35.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.35.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.36.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.36.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.36.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.36.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.36.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.36.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.36.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.36.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.36.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.36.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.36.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.36.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.36.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.36.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.36.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.36.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.36.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.36.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.36.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.36.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.36.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.37.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.37.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.37.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.37.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.37.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.37.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.37.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.37.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00054-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.37.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00075-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.37.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.37.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.37.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.37.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00093-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.37.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.37.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.37.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.37.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.37.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.37.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.37.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.37.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.38.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.38.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.38.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.38.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.38.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.38.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.38.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.38.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00107-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.38.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00055-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.38.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.38.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.38.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.38.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.38.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.38.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.38.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.38.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.38.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.38.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.38.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.38.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.39.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.39.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00062-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.39.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.39.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00029-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.39.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.39.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.39.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.39.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.39.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.39.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.39.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.39.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.39.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.39.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.39.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.39.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.39.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.39.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.39.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.39.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.39.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.4.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.4.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.4.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.4.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.4.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.4.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.4.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.4.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.4.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.4.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.4.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.4.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.4.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00029-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.4.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.4.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.4.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.4.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.4.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.4.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.4.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.4.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.40.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.40.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.40.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.40.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.40.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.40.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00097-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.40.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.40.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00055-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.40.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00102-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.40.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.40.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.40.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.40.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00086-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.40.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.40.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.40.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.40.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.40.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.40.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.40.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.40.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.41.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.41.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.41.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.41.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.41.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.41.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.41.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00060-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.41.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.41.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.41.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.41.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.41.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.41.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00030-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.41.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.41.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.41.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.41.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.41.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.41.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.41.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.41.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00051-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.42.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.42.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.42.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.42.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.42.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.42.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.42.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.42.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.42.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.42.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.42.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.42.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00035-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.42.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00091-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.42.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.42.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.42.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.42.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.42.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.42.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.42.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.42.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.43.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.43.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.43.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.43.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.43.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.43.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.43.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.43.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.43.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.43.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.43.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.43.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.43.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.43.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00047-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.43.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.43.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.43.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.43.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.43.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.43.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.43.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.44.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.44.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.44.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.44.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.44.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.44.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.44.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.44.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.44.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.44.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.44.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00097-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.44.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.44.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00005-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.44.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.44.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.44.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.44.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.44.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.44.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.44.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.44.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.45.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.45.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.45.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.45.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.45.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.45.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.45.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.45.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.45.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.45.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.45.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.45.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.45.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.45.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.45.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00005-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.45.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.45.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.45.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00097-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.45.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.45.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.45.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.46.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.46.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.46.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.46.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.46.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.46.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.46.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.46.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.46.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.46.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.46.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.46.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.46.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00008-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.46.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.46.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.46.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.46.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.46.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.46.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.46.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.46.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.47.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.47.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.47.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.47.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.47.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.47.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.47.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.47.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.47.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.47.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.47.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.47.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.47.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.47.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.47.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00066-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.47.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.47.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.47.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.47.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.47.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.47.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.48.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.48.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.48.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.48.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.48.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.48.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.48.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.48.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.48.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.48.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.48.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.48.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.48.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.48.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.48.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00075-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.48.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.48.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.48.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.48.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.48.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.48.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.49.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.49.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.49.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.49.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.49.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.49.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.49.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.49.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.49.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.49.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.49.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.49.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00088-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.49.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.49.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.49.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.49.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.49.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.49.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.49.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.49.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.49.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.5.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.5.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.5.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.5.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.5.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.5.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.5.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00015-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.5.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.5.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.5.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.5.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.5.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.5.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.5.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.5.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.5.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.5.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00086-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.5.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.5.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.5.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.5.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.50.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00086-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.50.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.50.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.50.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.50.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.50.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.50.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00051-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.50.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.50.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.50.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.50.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.50.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.50.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.50.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.50.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.50.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.50.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.50.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.50.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.50.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.50.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.51.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.51.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.51.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00033-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.51.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.51.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.51.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.51.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.51.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.51.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.51.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.51.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.51.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.51.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.51.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.51.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.51.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.51.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.51.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.51.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.51.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.51.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.52.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.52.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.52.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.52.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00086-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.52.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.52.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.52.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.52.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.52.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.52.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.52.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.52.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.52.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.52.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00075-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.52.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00022-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.52.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.52.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.52.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.52.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.52.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.52.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.53.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.53.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.53.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.53.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.53.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.53.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.53.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.53.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.53.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00030-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.53.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.53.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.53.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.53.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.53.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.53.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.53.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.53.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.53.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.53.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.53.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.53.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.54.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.54.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.54.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.54.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.54.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.54.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.54.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.54.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.54.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.54.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.54.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.54.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.54.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00012-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.54.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.54.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.54.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.54.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00080-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.54.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00007-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.54.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.54.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.54.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.55.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.55.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.55.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.55.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.55.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.55.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.55.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.55.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.55.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.55.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.55.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.55.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.55.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.55.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.55.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.55.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.55.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.55.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.55.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.55.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00077-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.55.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.56.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00011-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.56.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.56.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00100-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.56.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.56.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.56.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.56.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.56.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.56.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.56.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00093-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.56.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.56.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.56.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.56.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00001-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.56.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.56.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00038-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.56.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.56.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00105-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.56.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.56.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.56.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.57.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.57.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.57.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.57.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.57.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.57.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.57.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.57.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00017-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.57.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.57.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.57.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.57.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00012-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.57.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00062-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.57.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.57.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.57.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.57.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.57.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.57.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.57.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.57.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.58.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.58.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.58.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.58.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.58.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.58.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.58.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.58.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.58.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.58.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.58.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.58.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.58.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.58.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00064-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.58.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.58.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.58.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.58.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.58.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.58.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.58.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.59.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.59.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00075-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.59.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.59.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.59.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00040-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.59.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.59.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.59.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.59.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.59.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.59.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.59.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.59.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.59.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.59.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.59.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.59.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.59.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.59.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.59.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.59.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.6.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.6.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.6.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.6.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00012-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.6.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.6.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.6.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.6.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.6.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.6.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.6.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.6.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.6.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00101-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.6.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.6.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00073-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.6.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.6.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.6.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.6.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.6.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.6.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.60.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.60.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.60.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.60.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.60.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00051-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.60.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.60.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.60.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.60.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.60.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.60.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.60.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.60.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00074-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.60.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.60.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.60.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.60.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.60.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.60.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.60.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00058-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.60.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00049-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.61.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.61.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.61.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.61.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.61.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.61.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.61.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.61.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.61.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.61.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00078-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.61.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.61.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.61.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.61.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.61.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00051-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.61.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.61.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00060-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.61.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.61.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00038-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.61.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00023-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.61.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.62.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00025-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.62.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00013-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.62.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.62.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.62.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00047-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.62.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.62.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.62.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.62.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.62.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.62.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.62.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.62.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.62.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.62.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00071-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.62.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.62.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00085-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.62.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00087-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.62.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.62.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.62.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.63.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.63.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.63.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.63.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.63.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.63.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.63.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.63.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.63.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.63.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.63.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.63.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.63.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.63.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00019-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.63.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00092-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.63.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.63.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.63.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.63.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.63.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.63.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.64.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.64.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.64.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.64.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.64.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.64.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00079-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.64.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00004-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.64.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.64.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00106-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.64.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.64.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.64.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00072-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.64.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.64.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.64.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.64.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.64.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.64.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00098-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.64.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00028-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.64.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00053-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.64.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.65.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.65.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00084-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.65.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.65.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 32768,
"rule": "preserve",
"shape": [
16,
1024
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 32768
},
"model.llm.layers.65.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 8192,
"rule": "preserve",
"shape": [
1024,
1,
4
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 8192
},
"model.llm.layers.65.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.65.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.65.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00003-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.65.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.65.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.65.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00009-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.65.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00060-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.65.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00006-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.65.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.65.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.65.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.65.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.65.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.65.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.65.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.65.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.7.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00032-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.7.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00090-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.7.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00057-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.7.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00062-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.7.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.7.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.7.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.7.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.7.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.7.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00026-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.7.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00043-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.7.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.7.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.7.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00103-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.7.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00061-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.7.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.7.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.7.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00067-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.7.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00103-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.7.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00056-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.7.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.8.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.8.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.8.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.8.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00104-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.8.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00108-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.8.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.8.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00094-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.8.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00020-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.8.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.8.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.8.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00082-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.8.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00083-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.8.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00099-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.8.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.8.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00034-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.8.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00059-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.8.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.8.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.8.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00069-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.8.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00016-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.8.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.9.attn.k_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00042-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.9.attn.k_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00050-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.9.attn.q_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00081-of-00108.safetensors",
"source_bytes": 256
},
"model.llm.layers.9.attn.rel_logits_proj.proj": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
16,
512
],
"shard": "model-00024-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.9.attn.v_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 16384,
"rule": "preserve",
"shape": [
2048,
1,
4
],
"shard": "model-00046-of-00108.safetensors",
"source_bytes": 16384
},
"model.llm.layers.9.attn.wk_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00041-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.9.attn.wo_ud.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
6144,
8192
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.9.attn.wq_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 100663296,
"rule": "preserve",
"shape": [
8192,
6144
],
"shard": "model-00070-of-00108.safetensors",
"source_bytes": 100663296
},
"model.llm.layers.9.attn.wr_du.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12582912,
"rule": "preserve",
"shape": [
1024,
6144
],
"shard": "model-00036-of-00108.safetensors",
"source_bytes": 12582912
},
"model.llm.layers.9.attn.wv_dv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 25165824,
"rule": "preserve",
"shape": [
2048,
6144
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 25165824
},
"model.llm.layers.9.attn_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00052-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.9.attn_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.layers.9.mlp.experts.w13_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 150994944,
"scales": 150994944,
"total": 3925868544,
"weight": 3623878656
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 3925868544,
"rule": "base-routed-experts",
"shape": [
256,
6144,
6144
],
"shard": "model-00027-of-00108.safetensors",
"source_bytes": 19327352832
},
"model.llm.layers.9.mlp.experts.w2_weight": {
"action": "quantize",
"bits": 3,
"components": {
"biases": 75497472,
"scales": 75497472,
"total": 1962934272,
"weight": 1811939328
},
"dtype": "BF16",
"group_size": 128,
"mode": "affine",
"output_bytes": 1962934272,
"rule": "base-routed-experts",
"shape": [
256,
6144,
3072
],
"shard": "model-00048-of-00108.safetensors",
"source_bytes": 9663676416
},
"model.llm.layers.9.mlp.gate.bias": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 1024,
"rule": "preserve",
"shape": [
256
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 1024
},
"model.llm.layers.9.mlp.gate.global_scale": {
"action": "preserve",
"dtype": "F32",
"output_bytes": 4,
"rule": "preserve",
"shape": [
1
],
"shard": "model-00096-of-00108.safetensors",
"source_bytes": 4
},
"model.llm.layers.9.mlp.gate.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 3170304,
"rule": "preserve",
"shape": [
258,
6144
],
"shard": "model-00045-of-00108.safetensors",
"source_bytes": 3170304
},
"model.llm.layers.9.mlp.shared_experts.shared_w13_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 150994944,
"rule": "preserve",
"shape": [
2,
6144,
6144
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 150994944
},
"model.llm.layers.9.mlp.shared_experts.shared_w2_weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 75497472,
"rule": "preserve",
"shape": [
2,
6144,
3072
],
"shard": "model-00044-of-00108.safetensors",
"source_bytes": 75497472
},
"model.llm.layers.9.mlp_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.layers.9.mlp_sconv.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152,
"rule": "preserve",
"shape": [
6144,
1,
4
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 49152
},
"model.llm.norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00076-of-00108.safetensors",
"source_bytes": 12288
},
"model.llm.unembed.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 2470182912,
"rule": "preserve",
"shape": [
201024,
6144
],
"shard": "model-00089-of-00108.safetensors",
"source_bytes": 2470182912
},
"model.mtp.layers.0.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.0.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.0.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.0.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.0.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.0.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.0.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.0.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.0.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.0.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.0.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.0.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.0.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.0.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.0.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.0.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.0.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.0.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.0.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.0.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.1.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.1.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.1.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.1.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.1.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 8192
},
"model.mtp.layers.1.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.1.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
1024
],
"shard": "mtp.safetensors",
"source_bytes": 32768
},
"model.mtp.layers.1.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 8192
},
"model.mtp.layers.1.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.1.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.1.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.1.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.1.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.1.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.1.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.1.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.1.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.1.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.1.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.1.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.2.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.2.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.2.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.2.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.2.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.2.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.2.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.2.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.2.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.2.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.2.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.2.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.2.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.2.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.2.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.2.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.2.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.2.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.2.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.2.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.3.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.3.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.3.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.3.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.3.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 8192
},
"model.mtp.layers.3.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.3.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
1024
],
"shard": "mtp.safetensors",
"source_bytes": 32768
},
"model.mtp.layers.3.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 8192
},
"model.mtp.layers.3.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.3.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.3.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.3.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.3.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.3.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.3.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.3.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.3.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.3.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.3.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.3.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.4.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.4.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.4.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.4.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.4.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.4.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.4.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.4.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.4.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.4.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.4.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.4.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.4.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.4.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.4.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.4.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.4.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.4.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.4.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.4.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.5.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.5.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.5.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.5.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.5.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.5.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.5.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.5.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.5.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.5.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.5.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.5.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.5.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.5.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.5.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.5.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.5.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.5.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.5.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.5.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.6.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.6.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.6.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.6.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.6.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.6.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.6.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.6.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.6.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.6.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.6.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.6.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.6.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.6.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.6.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.6.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.6.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.6.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.6.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.6.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.7.embed_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.7.hidden_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.7.input_proj.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
12288
],
"shard": "mtp.safetensors",
"source_bytes": 150994944
},
"model.mtp.layers.7.transformer_block.attn.k_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.7.transformer_block.attn.k_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.7.transformer_block.attn.q_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
128
],
"shard": "mtp.safetensors",
"source_bytes": 256
},
"model.mtp.layers.7.transformer_block.attn.rel_logits_proj.proj": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
16,
512
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.7.transformer_block.attn.v_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 16384
},
"model.mtp.layers.7.transformer_block.attn.wk_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.7.transformer_block.attn.wo_ud.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
8192
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.7.transformer_block.attn.wq_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
8192,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 100663296
},
"model.mtp.layers.7.transformer_block.attn.wr_du.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1024,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12582912
},
"model.mtp.layers.7.transformer_block.attn.wv_dv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
2048,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 25165824
},
"model.mtp.layers.7.transformer_block.attn_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.7.transformer_block.attn_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.mtp.layers.7.transformer_block.mlp.global_scale": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
1
],
"shard": "mtp.safetensors",
"source_bytes": 2
},
"model.mtp.layers.7.transformer_block.mlp.w13_dn.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
49152,
6144
],
"shard": "mtp.safetensors",
"source_bytes": 603979776
},
"model.mtp.layers.7.transformer_block.mlp.w2_md.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
24576
],
"shard": "mtp.safetensors",
"source_bytes": 301989888
},
"model.mtp.layers.7.transformer_block.mlp_norm.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144
],
"shard": "mtp.safetensors",
"source_bytes": 12288
},
"model.mtp.layers.7.transformer_block.mlp_sconv.weight": {
"action": "drop",
"dtype": "BF16",
"output_bytes": 0,
"rule": "drop-prefix",
"shape": [
6144,
1,
4
],
"shard": "mtp.safetensors",
"source_bytes": 49152
},
"model.visual.final_norm.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 12288,
"rule": "preserve",
"shape": [
6144
],
"shard": "model-00002-of-00108.safetensors",
"source_bytes": 12288
},
"model.visual.layers.linear_0.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 19200,
"rule": "preserve",
"shape": [
128,
75
],
"shard": "model-00010-of-00108.safetensors",
"source_bytes": 19200
},
"model.visual.layers.linear_1.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 327680,
"rule": "preserve",
"shape": [
320,
512
],
"shard": "model-00031-of-00108.safetensors",
"source_bytes": 327680
},
"model.visual.layers.linear_2.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 49152000,
"rule": "preserve",
"shape": [
4800,
5120
],
"shard": "model-00037-of-00108.safetensors",
"source_bytes": 49152000
},
"model.visual.layers.linear_3.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 117964800,
"rule": "preserve",
"shape": [
6144,
9600
],
"shard": "model-00062-of-00108.safetensors",
"source_bytes": 117964800
},
"model.visual.layers.norm_0.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 256,
"rule": "preserve",
"shape": [
128
],
"shard": "model-00018-of-00108.safetensors",
"source_bytes": 256
},
"model.visual.layers.norm_1.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 640,
"rule": "preserve",
"shape": [
320
],
"shard": "model-00021-of-00108.safetensors",
"source_bytes": 640
},
"model.visual.layers.norm_2.weight": {
"action": "preserve",
"dtype": "BF16",
"output_bytes": 9600,
"rule": "preserve",
"shape": [
4800
],
"shard": "model-00039-of-00108.safetensors",
"source_bytes": 9600
}
},
"recipe": {
"description": "Preserve the official BF16 set and use a 3-bit/group-128 routed-expert baseline sized against the lower verified Metal working-set limit; promote only tensors selected by measured sensitivity.",
"drop_prefixes": [
"model.mtp."
],
"memory_budget_bytes": 483183820800,
"mode": "affine",
"name": "inkling-alis-quality-first",
"rules": [
{
"bits": 3,
"group_size": 128,
"name": "base-routed-experts",
"pattern": "^model\\.llm\\.layers\\.([3-9]|[1-5][0-9]|6[0-5])\\.mlp\\.experts\\.(w13_weight|w2_weight)$"
}
],
"runtime_reserve_floor_bytes": 34359738368,
"source_repo": "thinkingmachines/Inkling",
"source_revision": "e4aa5ee880fbb0d2c1a93b1e4a39f2d4b97eb28a"
},
"summary": {
"dropped_bytes": 10520891408,
"dropped_parameters": 5260445704,
"fits_static_budget": true,
"header_repo": "thinkingmachines/Inkling",
"header_revision": "e4aa5ee880fbb0d2c1a93b1e4a39f2d4b97eb28a",
"meets_reserve_floor": true,
"memory_budget_bytes": 483183820800,
"memory_budget_gib": 450.0,
"output_bytes": 438794123524,
"output_gib": 408.658872846514,
"output_logical_parameters": 947117177922,
"recipe": "inkling-alis-quality-first",
"rule_summary": {
"base-routed-experts": {
"output_bytes": 370994577408,
"source_bytes": 1826434842624,
"tensor_count": 126
},
"preserve": {
"output_bytes": 67799546116,
"source_bytes": 67799546116,
"tensor_count": 1266
}
},
"runtime_reserve_bytes": 44389697276,
"runtime_reserve_floor_bytes": 34359738368,
"runtime_reserve_gib": 41.34112715348601,
"source_bytes": 1904755280148,
"source_parameters": 952377623626,
"source_repo": "thinkingmachines/Inkling",
"source_revision": "e4aa5ee880fbb0d2c1a93b1e4a39f2d4b97eb28a",
"tensor_count": 1552
}
}