Files
Qwen3-4B-Thinking-2507-hqq-…/fake_quant_config.json
ModelHub XC 15e51f7507 初始化项目,由ModelHub XC社区提供模型
Model: meteorain/Qwen3-4B-Thinking-2507-hqq-fp4-e1m2-faked-bf16
Source: Original Platform
2026-09-16 04:02:17 +08:00

2545 lines
52 KiB
JSON

{
"method": "HQQ-style-FP4",
"format": "fp4-e1m2",
"nbits": 4,
"group_size": 128,
"axis": 1,
"round_zero": false,
"codebook": "fp4-e1m2",
"block_size": 128,
"scale_opt_steps": 8,
"optimize": true,
"optimize_device": "cuda",
"num_modules": 252,
"num_weights": 3633315840,
"elapsed_sec": 0.913,
"quantized_modules": [
{
"name": "model.layers.0.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.66658888803795e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.0.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.228014055726817e-06,
"max_abs_err": 0.0341796875
},
{
"name": "model.layers.0.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.497596132248873e-06,
"max_abs_err": 0.0234375
},
{
"name": "model.layers.0.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.845693522132933e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.0.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.682324510620674e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.0.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.8043774515681434e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.0.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.869829576316988e-06,
"max_abs_err": 0.099609375
},
{
"name": "model.layers.1.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 5.774431429017568e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.1.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.164545084175188e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.1.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.565248895640252e-06,
"max_abs_err": 0.0244140625
},
{
"name": "model.layers.1.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.783694632555125e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.1.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.26623614152777e-06,
"max_abs_err": 0.078125
},
{
"name": "model.layers.1.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 2.900868366850773e-06,
"max_abs_err": 0.0458984375
},
{
"name": "model.layers.1.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 3.430290234973654e-06,
"max_abs_err": 0.1328125
},
{
"name": "model.layers.2.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.1638543229491916e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.2.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.863240741949994e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.2.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.0754367698100396e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.2.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.978390163363656e-06,
"max_abs_err": 0.134765625
},
{
"name": "model.layers.2.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 4.402417289384175e-06,
"max_abs_err": 0.0537109375
},
{
"name": "model.layers.2.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 3.733332505362341e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.2.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.177442290005274e-06,
"max_abs_err": 0.10546875
},
{
"name": "model.layers.3.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.301024313870585e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.3.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.2831927607476246e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.3.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.215282039396698e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.3.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.041947926860303e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.3.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.628478670085315e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.3.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 4.191422249277821e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.3.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.108148343424546e-06,
"max_abs_err": 0.12890625
},
{
"name": "model.layers.4.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.47606429993175e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.4.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.877471150801284e-06,
"max_abs_err": 0.109375
},
{
"name": "model.layers.4.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.263982297445182e-06,
"max_abs_err": 0.02197265625
},
{
"name": "model.layers.4.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.268875040404964e-06,
"max_abs_err": 0.078125
},
{
"name": "model.layers.4.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 1.1854729564220179e-05,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.4.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.02237844557385e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.4.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.934885848546401e-06,
"max_abs_err": 0.09765625
},
{
"name": "model.layers.5.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.400415713869734e-06,
"max_abs_err": 0.0859375
},
{
"name": "model.layers.5.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.359171038639033e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.5.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.2252601057698485e-06,
"max_abs_err": 0.0263671875
},
{
"name": "model.layers.5.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.90084346185904e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.5.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 1.076699118129909e-05,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.5.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.640160907205427e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.5.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 5.952192623226438e-06,
"max_abs_err": 0.126953125
},
{
"name": "model.layers.6.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.841471531515708e-06,
"max_abs_err": 0.091796875
},
{
"name": "model.layers.6.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.355011919367826e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.6.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.466654551535612e-06,
"max_abs_err": 0.0478515625
},
{
"name": "model.layers.6.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.292298621701775e-06,
"max_abs_err": 0.107421875
},
{
"name": "model.layers.6.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 9.26383745536441e-06,
"max_abs_err": 0.0859375
},
{
"name": "model.layers.6.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.954497737548081e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.6.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.334091722237645e-06,
"max_abs_err": 0.109375
},
{
"name": "model.layers.7.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.405027761502424e-06,
"max_abs_err": 0.0595703125
},
{
"name": "model.layers.7.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.523011052195216e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.7.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.213954060920514e-06,
"max_abs_err": 0.02392578125
},
{
"name": "model.layers.7.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.175010639708489e-06,
"max_abs_err": 0.099609375
},
{
"name": "model.layers.7.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 9.049237633007579e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.7.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.994885214022361e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.7.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.178700459713582e-06,
"max_abs_err": 0.11328125
},
{
"name": "model.layers.8.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.4732412283774465e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.8.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.495269190054387e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.8.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.114209776977077e-06,
"max_abs_err": 0.02294921875
},
{
"name": "model.layers.8.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.415943971660454e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.8.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.722223017481156e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.8.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.2859485296939965e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.8.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.674546966678463e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.9.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.364603905240074e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.9.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.662680334557081e-06,
"max_abs_err": 0.09375
},
{
"name": "model.layers.9.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.579759196436498e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.9.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.3574275372957345e-06,
"max_abs_err": 0.140625
},
{
"name": "model.layers.9.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 9.045046681421809e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.9.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 5.79614197704359e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.9.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.035466867615469e-06,
"max_abs_err": 0.11328125
},
{
"name": "model.layers.10.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.551509613927919e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.10.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.589855274796719e-06,
"max_abs_err": 0.0595703125
},
{
"name": "model.layers.10.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.467636689601932e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.10.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.575722636625869e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.10.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 8.61238095239969e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.10.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.0256793403823394e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.10.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.35685046290746e-06,
"max_abs_err": 0.109375
},
{
"name": "model.layers.11.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.644221230089897e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.11.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.811941602791194e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.11.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.607318704889622e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.11.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.422683782147942e-06,
"max_abs_err": 0.11328125
},
{
"name": "model.layers.11.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 8.100550985545851e-06,
"max_abs_err": 0.091796875
},
{
"name": "model.layers.11.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.3916972976585384e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.11.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.57642067380948e-06,
"max_abs_err": 0.1484375
},
{
"name": "model.layers.12.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.720094916090602e-06,
"max_abs_err": 0.0634765625
},
{
"name": "model.layers.12.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.638781542278593e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.12.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.501674528815784e-06,
"max_abs_err": 0.02587890625
},
{
"name": "model.layers.12.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.356723588396562e-06,
"max_abs_err": 0.09765625
},
{
"name": "model.layers.12.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.770722731947899e-06,
"max_abs_err": 0.076171875
},
{
"name": "model.layers.12.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.74685634294292e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.12.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.801165454817237e-06,
"max_abs_err": 0.13671875
},
{
"name": "model.layers.13.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.749923613824649e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.13.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.736323030054336e-06,
"max_abs_err": 0.0859375
},
{
"name": "model.layers.13.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.98321889305953e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.13.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.071170446375618e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.13.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.328984338528244e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.13.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.985415893723257e-06,
"max_abs_err": 0.09375
},
{
"name": "model.layers.13.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.139888111851178e-06,
"max_abs_err": 0.171875
},
{
"name": "model.layers.14.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.594701972062467e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.14.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.503130407509161e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.14.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.752906640234869e-06,
"max_abs_err": 0.03076171875
},
{
"name": "model.layers.14.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.896986749576172e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.14.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.1082763497543056e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.14.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.806223154853797e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.14.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.982676040934166e-06,
"max_abs_err": 0.1953125
},
{
"name": "model.layers.15.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.793315151298884e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.15.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.60156229059794e-06,
"max_abs_err": 0.0576171875
},
{
"name": "model.layers.15.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.4825934461841825e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.15.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.58813508177991e-06,
"max_abs_err": 0.0478515625
},
{
"name": "model.layers.15.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.681725153612206e-06,
"max_abs_err": 0.107421875
},
{
"name": "model.layers.15.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.67704898660304e-06,
"max_abs_err": 0.0615234375
},
{
"name": "model.layers.15.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.025957074802136e-06,
"max_abs_err": 0.12109375
},
{
"name": "model.layers.16.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.375475095410366e-06,
"max_abs_err": 0.0458984375
},
{
"name": "model.layers.16.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 6.075618784961989e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.16.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.547460882051382e-06,
"max_abs_err": 0.02734375
},
{
"name": "model.layers.16.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.888903615414165e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.16.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.974292773520574e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.16.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.592288173123961e-06,
"max_abs_err": 0.14453125
},
{
"name": "model.layers.16.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.719805242028087e-06,
"max_abs_err": 0.15234375
},
{
"name": "model.layers.17.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.713309176120674e-06,
"max_abs_err": 0.0517578125
},
{
"name": "model.layers.17.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.78036269871518e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.17.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.213300705188885e-06,
"max_abs_err": 0.0302734375
},
{
"name": "model.layers.17.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.717783551517641e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.17.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.788160590076586e-06,
"max_abs_err": 0.087890625
},
{
"name": "model.layers.17.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.571437097591115e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.17.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.798643880756572e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.18.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.246254088182468e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.18.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.814269570691977e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.18.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.3676415013324e-06,
"max_abs_err": 0.021484375
},
{
"name": "model.layers.18.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.691366368409945e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.18.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.704062343487749e-06,
"max_abs_err": 0.0712890625
},
{
"name": "model.layers.18.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.634804776695091e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.18.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.876002771605272e-06,
"max_abs_err": 0.09765625
},
{
"name": "model.layers.19.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.555923846462974e-06,
"max_abs_err": 0.0517578125
},
{
"name": "model.layers.19.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.588254680333193e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.19.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.1156982812681235e-06,
"max_abs_err": 0.02734375
},
{
"name": "model.layers.19.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.875853730685776e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.19.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.645252142334357e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.19.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.691770522593288e-06,
"max_abs_err": 0.06201171875
},
{
"name": "model.layers.19.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.929201845196076e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.20.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.956498054933036e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.20.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.5764585340511985e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.20.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.1822469180915505e-06,
"max_abs_err": 0.02783203125
},
{
"name": "model.layers.20.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.533793228096329e-06,
"max_abs_err": 0.052734375
},
{
"name": "model.layers.20.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.440424385800725e-06,
"max_abs_err": 0.083984375
},
{
"name": "model.layers.20.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.626226422667969e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.20.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.890274107718142e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.21.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.687082532152999e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.21.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.4231272770266514e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.21.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.0408468673122115e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.21.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.424541996035259e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.21.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.308795491349883e-06,
"max_abs_err": 0.1171875
},
{
"name": "model.layers.21.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.573054633918218e-06,
"max_abs_err": 0.0556640625
},
{
"name": "model.layers.21.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.9304323915275745e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.22.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.425628271244932e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.22.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.338008122635074e-06,
"max_abs_err": 0.09765625
},
{
"name": "model.layers.22.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.1869012572278734e-06,
"max_abs_err": 0.0400390625
},
{
"name": "model.layers.22.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.638883976644138e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.22.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.7656374085345306e-06,
"max_abs_err": 0.060546875
},
{
"name": "model.layers.22.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.894521447975421e-06,
"max_abs_err": 0.06396484375
},
{
"name": "model.layers.22.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.1611625571676996e-06,
"max_abs_err": 0.07421875
},
{
"name": "model.layers.23.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 7.380164788628463e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.23.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.645881628879579e-06,
"max_abs_err": 0.0419921875
},
{
"name": "model.layers.23.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.65255117585184e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.23.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.053089691704372e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.23.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.1770637077861466e-06,
"max_abs_err": 0.087890625
},
{
"name": "model.layers.23.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.846154974482488e-06,
"max_abs_err": 0.0537109375
},
{
"name": "model.layers.23.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.018532869551564e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.24.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 7.523407020926243e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.24.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.696314929082291e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.24.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.617234132339945e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.24.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.203766133694444e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.24.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.4055001277884e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.24.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.851425951026613e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.24.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.915386620676145e-06,
"max_abs_err": 0.08984375
},
{
"name": "model.layers.25.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.528834546770668e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.25.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.826705091749318e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.25.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.601931429235265e-06,
"max_abs_err": 0.025390625
},
{
"name": "model.layers.25.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.936447450949345e-06,
"max_abs_err": 0.02685546875
},
{
"name": "model.layers.25.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.535476925113471e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.25.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.8413132794376e-06,
"max_abs_err": 0.056640625
},
{
"name": "model.layers.25.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.9217171585478354e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.26.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.854362709418638e-06,
"max_abs_err": 0.0537109375
},
{
"name": "model.layers.26.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.725560185965151e-06,
"max_abs_err": 0.04296875
},
{
"name": "model.layers.26.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.633677341800649e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.26.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.25323173153447e-06,
"max_abs_err": 0.033203125
},
{
"name": "model.layers.26.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.4908325586875435e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.26.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.945778295630589e-06,
"max_abs_err": 0.0556640625
},
{
"name": "model.layers.26.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 6.876712177472655e-06,
"max_abs_err": 0.083984375
},
{
"name": "model.layers.27.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.604020654776832e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.27.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.270923338684952e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.27.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.308172826014925e-06,
"max_abs_err": 0.0361328125
},
{
"name": "model.layers.27.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.005186151014641e-06,
"max_abs_err": 0.037109375
},
{
"name": "model.layers.27.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.4246859185223e-06,
"max_abs_err": 0.064453125
},
{
"name": "model.layers.27.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.1098124863056e-06,
"max_abs_err": 0.0439453125
},
{
"name": "model.layers.27.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.034323971311096e-06,
"max_abs_err": 0.072265625
},
{
"name": "model.layers.28.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.822126579208998e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.28.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.610631433228264e-06,
"max_abs_err": 0.0380859375
},
{
"name": "model.layers.28.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.827448825992178e-06,
"max_abs_err": 0.0283203125
},
{
"name": "model.layers.28.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 5.911361313337693e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.28.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.272713446582202e-06,
"max_abs_err": 0.078125
},
{
"name": "model.layers.28.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.251035185618093e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.28.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.172786808951059e-06,
"max_abs_err": 0.1171875
},
{
"name": "model.layers.29.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 7.654311957594473e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.29.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.562735168496147e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.29.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 7.4582653724064585e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.29.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.384533662640024e-06,
"max_abs_err": 0.029296875
},
{
"name": "model.layers.29.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.19490208211937e-06,
"max_abs_err": 0.0703125
},
{
"name": "model.layers.29.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.442420610459521e-06,
"max_abs_err": 0.0517578125
},
{
"name": "model.layers.29.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.223735337902326e-06,
"max_abs_err": 0.06640625
},
{
"name": "model.layers.30.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.460778877226403e-06,
"max_abs_err": 0.041015625
},
{
"name": "model.layers.30.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.865774255653378e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.30.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.244615855801385e-06,
"max_abs_err": 0.0322265625
},
{
"name": "model.layers.30.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.266686341405148e-06,
"max_abs_err": 0.0302734375
},
{
"name": "model.layers.30.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.119496331142727e-06,
"max_abs_err": 0.1015625
},
{
"name": "model.layers.30.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.610458396811737e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.30.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.440278750436846e-06,
"max_abs_err": 0.087890625
},
{
"name": "model.layers.31.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.716814368701307e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.31.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.503860393218929e-06,
"max_abs_err": 0.044921875
},
{
"name": "model.layers.31.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.871877980709542e-06,
"max_abs_err": 0.03125
},
{
"name": "model.layers.31.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.682083949272055e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.31.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.9706256908830255e-06,
"max_abs_err": 0.095703125
},
{
"name": "model.layers.31.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.812638614268508e-06,
"max_abs_err": 0.0625
},
{
"name": "model.layers.31.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.704159543209244e-06,
"max_abs_err": 0.0859375
},
{
"name": "model.layers.32.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.567046511918306e-06,
"max_abs_err": 0.0478515625
},
{
"name": "model.layers.32.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.513081305252854e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.32.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.968218935478944e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.32.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.652267529716482e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.32.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.837972705397988e-06,
"max_abs_err": 0.10546875
},
{
"name": "model.layers.32.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.872768946981523e-06,
"max_abs_err": 0.068359375
},
{
"name": "model.layers.32.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.74850559537299e-06,
"max_abs_err": 0.09375
},
{
"name": "model.layers.33.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.746202416252345e-06,
"max_abs_err": 0.0576171875
},
{
"name": "model.layers.33.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 4.544934199657291e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.33.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 9.995132131734863e-06,
"max_abs_err": 0.0498046875
},
{
"name": "model.layers.33.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 7.230280516523635e-06,
"max_abs_err": 0.05859375
},
{
"name": "model.layers.33.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 6.776806912967004e-06,
"max_abs_err": 0.11328125
},
{
"name": "model.layers.33.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.96057065599598e-06,
"max_abs_err": 0.080078125
},
{
"name": "model.layers.33.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.64236847317079e-06,
"max_abs_err": 0.09375
},
{
"name": "model.layers.34.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.169498647068394e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.34.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.10346171722631e-06,
"max_abs_err": 0.0390625
},
{
"name": "model.layers.34.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 9.921202035911847e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.34.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.900201242387993e-06,
"max_abs_err": 0.0546875
},
{
"name": "model.layers.34.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.031401310086949e-06,
"max_abs_err": 0.08203125
},
{
"name": "model.layers.34.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.764217116346117e-06,
"max_abs_err": 0.07421875
},
{
"name": "model.layers.34.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.15795158612309e-06,
"max_abs_err": 0.125
},
{
"name": "model.layers.35.self_attn.q_proj",
"shape": [
4096,
2560
],
"dtype": "bfloat16",
"mse": 6.106618002377218e-06,
"max_abs_err": 0.05078125
},
{
"name": "model.layers.35.self_attn.k_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 5.7398196986468975e-06,
"max_abs_err": 0.048828125
},
{
"name": "model.layers.35.self_attn.v_proj",
"shape": [
1024,
2560
],
"dtype": "bfloat16",
"mse": 8.653964869154152e-06,
"max_abs_err": 0.03515625
},
{
"name": "model.layers.35.self_attn.o_proj",
"shape": [
2560,
4096
],
"dtype": "bfloat16",
"mse": 6.472256700362777e-06,
"max_abs_err": 0.046875
},
{
"name": "model.layers.35.mlp.gate_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 7.728675882390235e-06,
"max_abs_err": 0.11328125
},
{
"name": "model.layers.35.mlp.up_proj",
"shape": [
9728,
2560
],
"dtype": "bfloat16",
"mse": 8.014725608518347e-06,
"max_abs_err": 0.16015625
},
{
"name": "model.layers.35.mlp.down_proj",
"shape": [
2560,
9728
],
"dtype": "bfloat16",
"mse": 7.0698133640689775e-06,
"max_abs_err": 0.10546875
}
],
"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."
}