Files
Qwen3-4B-Thinking-2507-hqq-…/fake_quant_config.json
ModelHub XC 3f4dc694e9 初始化项目,由ModelHub XC社区提供模型
Model: meteorain/Qwen3-4B-Thinking-2507-hqq-fp4-e2m1-faked-bf16
Source: Original Platform
2026-08-28 19:58:32 +08:00

2545 lines
52 KiB
JSON

{
"method": "HQQ-style-FP4",
"format": "fp4-e2m1",
"nbits": 4,
"group_size": 128,
"axis": 1,
"round_zero": false,
"codebook": "fp4-e2m1",
"block_size": 128,
"scale_opt_steps": 8,
"optimize": true,
"optimize_device": "cuda",
"num_modules": 252,
"num_weights": 3633315840,
"elapsed_sec": 0.915,
"quantized_modules": [
{
"name": "model.layers.0.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.723196409235243e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.0.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.307804142124951e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.0.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.5842610890977085e-06,
"max_abs_err": 0.0185546875
},
{
"name": "model.layers.0.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.00482929055579e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.0.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.879803211428225e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.0.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.118308763485402e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.0.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.772256827185629e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.1.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 4.959922989655752e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.1.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.345012596080778e-06,
"max_abs_err": 0.02490234375
},
{
"name": "model.layers.1.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.70992915527313e-06,
"max_abs_err": 0.02001953125
},
{
"name": "model.layers.1.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.077310106571531e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.1.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.5506566241092514e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.1.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 2.523945340726641e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.1.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 2.9067582545394544e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.2.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.2966347539040726e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.2.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.751754997618264e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.2.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.152061814646004e-06,
"max_abs_err": 0.017578125
},
{
"name": "model.layers.2.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.296973085933132e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.2.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 3.7920920021861093e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.2.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 3.1710897019365802e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.2.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 3.5821180972561706e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.3.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.346242232917575e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.3.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.364207936509047e-06,
"max_abs_err": 0.03125
},
{
"name": "model.layers.3.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.239326012291713e-06,
"max_abs_err": 0.01904296875
},
{
"name": "model.layers.3.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.366312052501598e-06,
"max_abs_err": 0.0595703125
},
{
"name": "model.layers.3.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.701893653371371e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.3.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 3.6226672364136903e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.3.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 4.026178885396803e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.4.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.538647201319691e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.4.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.714638973586261e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.4.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.348350780172041e-06,
"max_abs_err": 0.0185546875
},
{
"name": "model.layers.4.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.599338237516349e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.4.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 1.0382831533206627e-05,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.4.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 4.3855334297404625e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.4.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 4.6982522690086626e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.5.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.475285433931276e-06,
"max_abs_err": 0.0458984375
},
{
"name": "model.layers.5.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.448680440167664e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.5.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.26765177003108e-06,
"max_abs_err": 0.021484375
},
{
"name": "model.layers.5.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.2222653721401e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.5.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 9.155895895673893e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.5.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 4.926101155433571e-06,
"max_abs_err": 0.0341796875
},
{
"name": "model.layers.5.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 4.911782525596209e-06,
"max_abs_err": 0.083984375
},
{
"name": "model.layers.6.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.81708764002542e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.6.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.4791053116787225e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.6.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.326151833491167e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.6.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.582049197982997e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.6.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 8.062746928771958e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.6.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.2111777222307865e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.6.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.27932525073993e-06,
"max_abs_err": 0.0859375
},
{
"name": "model.layers.7.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.479767423821613e-06,
"max_abs_err": 0.0478515625
},
{
"name": "model.layers.7.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.5348955356748775e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.7.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.168553111114306e-06,
"max_abs_err": 0.0185546875
},
{
"name": "model.layers.7.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.450538537843386e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.7.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.879278200562112e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.7.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.269621851766715e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.7.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.28146938449936e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.8.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.5435507420042995e-06,
"max_abs_err": 0.0341796875
},
{
"name": "model.layers.8.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.646627414535033e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.8.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.11546044840361e-06,
"max_abs_err": 0.017578125
},
{
"name": "model.layers.8.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.689119461749215e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.8.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.817412668169709e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.8.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.5131222325144336e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.8.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.620345291390549e-06,
"max_abs_err": 0.13671875
},
{
"name": "model.layers.9.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.33408683622838e-06,
"max_abs_err": 0.0478515625
},
{
"name": "model.layers.9.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.428300482890336e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.9.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.340322099480545e-06,
"max_abs_err": 0.0205078125
},
{
"name": "model.layers.9.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.603774297924247e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.9.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.977320819918532e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.9.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.0720645958790556e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.9.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.048508228355786e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.10.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.606837476079818e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.10.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.607840648735873e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.10.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.206769168988103e-06,
"max_abs_err": 0.0234375
},
{
"name": "model.layers.10.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.806037279398879e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.10.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.620466476510046e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.10.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.274556770018535e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.10.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.311468157742638e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.11.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.624520326819038e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.11.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.829042947880225e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.11.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.359824849118013e-06,
"max_abs_err": 0.01953125
},
{
"name": "model.layers.11.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.6785102060530335e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.11.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.147060387069359e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.11.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.590485670836642e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.11.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.537239303521346e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.12.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.72002454646281e-06,
"max_abs_err": 0.0458984375
},
{
"name": "model.layers.12.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.604595116892597e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.12.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.22695358490455e-06,
"max_abs_err": 0.02294921875
},
{
"name": "model.layers.12.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.586750376096461e-06,
"max_abs_err": 0.1015625
},
{
"name": "model.layers.12.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.797561582061462e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.12.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.845269242854556e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.12.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.679354217136279e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.13.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.685677933797706e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.13.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.608666015177732e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.13.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.8988301791250706e-06,
"max_abs_err": 0.0224609375
},
{
"name": "model.layers.13.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.32990088686347e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.13.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.356177436828148e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.13.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.0138231674500275e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.13.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.799479367851745e-06,
"max_abs_err": 0.109375
},
{
"name": "model.layers.14.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.6494714044674765e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.14.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.58950478080078e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.14.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.566788215423003e-06,
"max_abs_err": 0.01904296875
},
{
"name": "model.layers.14.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.186265752854524e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.14.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.17779141975916e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.14.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.8456639635551255e-06,
"max_abs_err": 0.0556640625
},
{
"name": "model.layers.14.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.66817152503063e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.15.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.792805495730136e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.15.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.5728778534103185e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.15.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.4250693867506925e-06,
"max_abs_err": 0.0244140625
},
{
"name": "model.layers.15.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 4.887425802735379e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.15.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.764145043940516e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.15.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.721125944546657e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.15.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.613454959529918e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.16.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.454677648231154e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.16.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.204238277656259e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.16.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.4383107201138046e-06,
"max_abs_err": 0.0205078125
},
{
"name": "model.layers.16.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.174977104616119e-06,
"max_abs_err": 0.07421875
},
{
"name": "model.layers.16.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.043761004548287e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.16.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.647351827064995e-06,
"max_abs_err": 0.099609375
},
{
"name": "model.layers.16.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.415483428805601e-06,
"max_abs_err": 0.140625
},
{
"name": "model.layers.17.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.737740139011294e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.17.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.916878424410243e-06,
"max_abs_err": 0.0341796875
},
{
"name": "model.layers.17.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.139554898254573e-06,
"max_abs_err": 0.0244140625
},
{
"name": "model.layers.17.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.00591659147176e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.17.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.881236575078219e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.17.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.679994501406327e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.17.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.455374321172712e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.18.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.35116441824357e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.18.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.0058010856446344e-06,
"max_abs_err": 0.0302734375
},
{
"name": "model.layers.18.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.356754056469072e-06,
"max_abs_err": 0.02001953125
},
{
"name": "model.layers.18.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.0082362577086315e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.18.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.761946795246331e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.18.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.685592441295739e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.18.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.464952209877083e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.19.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.572399913944537e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.19.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.739893029181985e-06,
"max_abs_err": 0.0537109375
},
{
"name": "model.layers.19.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.0880388446094e-06,
"max_abs_err": 0.02099609375
},
{
"name": "model.layers.19.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.105246145831188e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.19.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.692868398909923e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.19.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.7287729759991635e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.19.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.483118002302945e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.20.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.9451963352330495e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.20.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.642794465326006e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.20.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.088903774070786e-06,
"max_abs_err": 0.0234375
},
{
"name": "model.layers.20.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 4.850856839766493e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.20.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.563155355048366e-06,
"max_abs_err": 0.0595703125
},
{
"name": "model.layers.20.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.724029051634716e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.20.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.532818704523379e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.21.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.687680186383659e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.21.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.584234829962952e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.21.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.92296373724821e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.21.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 4.720107881439617e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.21.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.400830104917986e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.21.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.686766598955728e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.21.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.59258251087158e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.22.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.519015303434571e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.22.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.358698788564652e-06,
"max_abs_err": 0.087890625
},
{
"name": "model.layers.22.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.932849944656482e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.22.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 4.918535978504224e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.22.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.786592737422325e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.22.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.962916930002393e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.22.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.697096639778465e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.23.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.271174697758397e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.23.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.773907221533591e-06,
"max_abs_err": 0.0341796875
},
{
"name": "model.layers.23.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.224790013220627e-06,
"max_abs_err": 0.02685546875
},
{
"name": "model.layers.23.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.282028268993599e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.23.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.191915417730343e-06,
"max_abs_err": 0.07421875
},
{
"name": "model.layers.23.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.976956344966311e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.23.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.718308784707915e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.24.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.378887519531418e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.24.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.769645329361083e-06,
"max_abs_err": 0.02783203125
},
{
"name": "model.layers.24.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.1439882301783655e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.24.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.45633747606189e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.24.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.41799897493911e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.24.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.007910542393802e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.24.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.764070920122322e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.25.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.6406238400086295e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.25.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.04299578096834e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.25.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.485302037617657e-06,
"max_abs_err": 0.02490234375
},
{
"name": "model.layers.25.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.2846712605969515e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.25.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.560270776390098e-06,
"max_abs_err": 0.0517578125
},
{
"name": "model.layers.25.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.010762263031211e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.25.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.803109161206521e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.26.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.87312661082251e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.26.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.9139866860059556e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.26.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.441806363000069e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.26.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.529534519155277e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.26.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.563839178852504e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.26.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.136965112091275e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.26.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.901018994336482e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.27.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.742789198848186e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.27.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.546225682133809e-06,
"max_abs_err": 0.0244140625
},
{
"name": "model.layers.27.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.108388333814219e-06,
"max_abs_err": 0.02587890625
},
{
"name": "model.layers.27.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.291121397021925e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.27.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.50322226647404e-06,
"max_abs_err": 0.0556640625
},
{
"name": "model.layers.27.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.284828486968763e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.27.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.042769200576004e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.28.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.973874976916704e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.28.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.845411240239628e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.28.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.6871652961708605e-06,
"max_abs_err": 0.0205078125
},
{
"name": "model.layers.28.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.248173692962155e-06,
"max_abs_err": 0.0458984375
},
{
"name": "model.layers.28.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.368768026732141e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.28.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.419048077077605e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.28.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.186238351801876e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.29.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.596846105821896e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.29.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.675573563872604e-06,
"max_abs_err": 0.03125
},
{
"name": "model.layers.29.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.202787517395336e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.29.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.656474513671128e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.29.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.354502602334833e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.29.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.6106676968047395e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.29.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.352224318106892e-06,
"max_abs_err": 0.0556640625
},
{
"name": "model.layers.30.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.563273589359596e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.30.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.024985512136482e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.30.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.922944066900527e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.30.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.561913894780446e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.30.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.283879883994814e-06,
"max_abs_err": 0.07421875
},
{
"name": "model.layers.30.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.767488684999989e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.30.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.518580448755529e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.31.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.785035227745539e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.31.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.665187589125708e-06,
"max_abs_err": 0.03125
},
{
"name": "model.layers.31.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.459275366272777e-06,
"max_abs_err": 0.02734375
},
{
"name": "model.layers.31.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.895470621908316e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.31.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.151527941256063e-06,
"max_abs_err": 0.1015625
},
{
"name": "model.layers.31.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.928547009010799e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.31.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.769980245735496e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.32.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.622670414595632e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.32.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.707270818471443e-06,
"max_abs_err": 0.03125
},
{
"name": "model.layers.32.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.457750598405255e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.32.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.869081633136375e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.32.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.986003088764846e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.32.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.971795755816856e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.32.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.791136911488138e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.33.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.7098836805380415e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.33.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 3.815228865278186e-06,
"max_abs_err": 0.0263671875
},
{
"name": "model.layers.33.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.46271632326534e-06,
"max_abs_err": 0.0263671875
},
{
"name": "model.layers.33.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.374186341417953e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.33.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.882290224690223e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.33.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.030158485576976e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.33.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.649414444837021e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.34.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.292181867844192e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.34.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.395277755975258e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.34.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.383262866118457e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.34.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.0706788644893095e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.34.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.111520178819774e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.34.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.85034729031031e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.34.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.019272859703051e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.35.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.363972832128638e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.35.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.986941803508671e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.35.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.390552582364762e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.35.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.687425527867163e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.35.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.747232419002103e-06,
"max_abs_err": 0.109375
},
{
"name": "model.layers.35.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.060730695229722e-06,
"max_abs_err": 0.078125
},
{
"name": "model.layers.35.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.596617938863346e-06,
"max_abs_err": 0.07421875
}
],
"source_model": "Qwen/Qwen3-4B-Thinking-2507",
"dtype": "bf16",
"target_regex": "(?:q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)$",
"ignore_regex": [],
"include_lm_head": false,
"note": "Dense fake quant: quantized weights are dequantized and saved as normal HF tensors. Codebook formats are HQQ-style optimized fake-quant variants, not packed runtime checkpoints."
}