{ "format": "int3-sym", "codebook": "int3-sym", "block_size": 128, "num_modules": 252, "num_weights": 3633315840, "elapsed_sec": 92.489, "quantized_modules": [ { "name": "model.layers.0.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.120008088648319e-05, "max_abs_err": 0.09423828125 }, { "name": "model.layers.0.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.4661632273346186e-05, "max_abs_err": 0.048095703125 }, { "name": "model.layers.0.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.0131031710188836e-05, "max_abs_err": 0.027587890625 }, { "name": "model.layers.0.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.598972034524195e-05, "max_abs_err": 0.0849609375 }, { "name": "model.layers.0.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.095189433428459e-05, "max_abs_err": 0.0693359375 }, { "name": "model.layers.0.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.5615441447589546e-05, "max_abs_err": 0.0537109375 }, { "name": "model.layers.0.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.2330979340476915e-05, "max_abs_err": 0.08984375 }, { "name": "model.layers.1.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.562826896086335e-05, "max_abs_err": 0.061279296875 }, { "name": "model.layers.1.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.801125058089383e-05, "max_abs_err": 0.0380859375 }, { "name": "model.layers.1.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.02810619561933e-05, "max_abs_err": 0.02392578125 }, { "name": "model.layers.1.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.569530963432044e-05, "max_abs_err": 0.06689453125 }, { "name": "model.layers.1.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.7905778299318627e-05, "max_abs_err": 0.064453125 }, { "name": "model.layers.1.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 1.7861950254882686e-05, "max_abs_err": 0.049072265625 }, { "name": "model.layers.1.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 2.0824900275329128e-05, "max_abs_err": 0.11767578125 }, { "name": "model.layers.2.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.7999860069248825e-05, "max_abs_err": 0.05517578125 }, { "name": "model.layers.2.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.18409108533524e-05, "max_abs_err": 0.0625 }, { "name": "model.layers.2.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.353176700533368e-05, "max_abs_err": 0.0361328125 }, { "name": "model.layers.2.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.690033918246627e-05, "max_abs_err": 0.09375 }, { "name": "model.layers.2.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 2.6607118343235925e-05, "max_abs_err": 0.05029296875 }, { "name": "model.layers.2.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 2.298475374118425e-05, "max_abs_err": 0.0791015625 }, { "name": "model.layers.2.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 2.341487561352551e-05, "max_abs_err": 0.1279296875 }, { "name": "model.layers.3.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.881484008161351e-05, "max_abs_err": 0.07275390625 }, { "name": "model.layers.3.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.857341653201729e-05, "max_abs_err": 0.056884765625 }, { "name": "model.layers.3.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.440254269866273e-05, "max_abs_err": 0.031005859375 }, { "name": "model.layers.3.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.730829484993592e-05, "max_abs_err": 0.0673828125 }, { "name": "model.layers.3.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.656468809116632e-05, "max_abs_err": 0.07470703125 }, { "name": "model.layers.3.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 2.578065323177725e-05, "max_abs_err": 0.0732421875 }, { "name": "model.layers.3.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.081412069150247e-05, "max_abs_err": 0.11767578125 }, { "name": "model.layers.4.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.998504689661786e-05, "max_abs_err": 0.06982421875 }, { "name": "model.layers.4.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.182901466265321e-05, "max_abs_err": 0.12109375 }, { "name": "model.layers.4.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.4716311094816774e-05, "max_abs_err": 0.0308837890625 }, { "name": "model.layers.4.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.8662201404804364e-05, "max_abs_err": 0.0791015625 }, { "name": "model.layers.4.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 7.315659604500979e-05, "max_abs_err": 0.05126953125 }, { "name": "model.layers.4.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.08669768855907e-05, "max_abs_err": 0.059326171875 }, { "name": "model.layers.4.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.57640310539864e-05, "max_abs_err": 0.10693359375 }, { "name": "model.layers.5.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.9498740079579875e-05, "max_abs_err": 0.0732421875 }, { "name": "model.layers.5.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.909489169018343e-05, "max_abs_err": 0.055419921875 }, { "name": "model.layers.5.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.4471235014498234e-05, "max_abs_err": 0.0277099609375 }, { "name": "model.layers.5.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.640407885541208e-05, "max_abs_err": 0.06982421875 }, { "name": "model.layers.5.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 6.626145477639511e-05, "max_abs_err": 0.0791015625 }, { "name": "model.layers.5.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.4645494451979175e-05, "max_abs_err": 0.09326171875 }, { "name": "model.layers.5.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.644406751845963e-05, "max_abs_err": 0.11279296875 }, { "name": "model.layers.6.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.1985280404333025e-05, "max_abs_err": 0.10205078125 }, { "name": "model.layers.6.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.8988822780083865e-05, "max_abs_err": 0.060546875 }, { "name": "model.layers.6.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.57107053080108e-05, "max_abs_err": 0.042236328125 }, { "name": "model.layers.6.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.875873881042935e-05, "max_abs_err": 0.08642578125 }, { "name": "model.layers.6.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 5.709498509531841e-05, "max_abs_err": 0.0771484375 }, { "name": "model.layers.6.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.6569956137100235e-05, "max_abs_err": 0.080078125 }, { "name": "model.layers.6.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.866213228320703e-05, "max_abs_err": 0.126953125 }, { "name": "model.layers.7.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.9545098843518645e-05, "max_abs_err": 0.062255859375 }, { "name": "model.layers.7.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.960570757044479e-05, "max_abs_err": 0.0732421875 }, { "name": "model.layers.7.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.071053237770684e-05, "max_abs_err": 0.0255126953125 }, { "name": "model.layers.7.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.809275949606672e-05, "max_abs_err": 0.09765625 }, { "name": "model.layers.7.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 5.575290197157301e-05, "max_abs_err": 0.08203125 }, { "name": "model.layers.7.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.686954005388543e-05, "max_abs_err": 0.0791015625 }, { "name": "model.layers.7.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.800027116085403e-05, "max_abs_err": 0.1513671875 }, { "name": "model.layers.8.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.9973594539333135e-05, "max_abs_err": 0.05859375 }, { "name": "model.layers.8.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.004590300610289e-05, "max_abs_err": 0.057373046875 }, { "name": "model.layers.8.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.001741737942211e-05, "max_abs_err": 0.023681640625 }, { "name": "model.layers.8.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.954902058467269e-05, "max_abs_err": 0.0751953125 }, { "name": "model.layers.8.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.7601686674170196e-05, "max_abs_err": 0.08203125 }, { "name": "model.layers.8.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.8703143218299374e-05, "max_abs_err": 0.087890625 }, { "name": "model.layers.8.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.114795956411399e-05, "max_abs_err": 0.1513671875 }, { "name": "model.layers.9.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.917376307072118e-05, "max_abs_err": 0.091796875 }, { "name": "model.layers.9.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.985212242696434e-05, "max_abs_err": 0.12451171875 }, { "name": "model.layers.9.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.284209692035802e-05, "max_abs_err": 0.026123046875 }, { "name": "model.layers.9.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.925716009689495e-05, "max_abs_err": 0.126953125 }, { "name": "model.layers.9.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 5.5754433560650796e-05, "max_abs_err": 0.078125 }, { "name": "model.layers.9.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.569008913473226e-05, "max_abs_err": 0.0703125 }, { "name": "model.layers.9.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.721167013281956e-05, "max_abs_err": 0.10791015625 }, { "name": "model.layers.10.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.0449573134537786e-05, "max_abs_err": 0.052978515625 }, { "name": "model.layers.10.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.056708712596446e-05, "max_abs_err": 0.0771484375 }, { "name": "model.layers.10.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.232503099250607e-05, "max_abs_err": 0.03125 }, { "name": "model.layers.10.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 4.0596143662696704e-05, "max_abs_err": 0.08984375 }, { "name": "model.layers.10.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 5.30084507772699e-05, "max_abs_err": 0.09814453125 }, { "name": "model.layers.10.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.711859608301893e-05, "max_abs_err": 0.08056640625 }, { "name": "model.layers.10.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 3.917381400242448e-05, "max_abs_err": 0.166015625 }, { "name": "model.layers.11.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.093761162948795e-05, "max_abs_err": 0.061279296875 }, { "name": "model.layers.11.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.141444151173346e-05, "max_abs_err": 0.06005859375 }, { "name": "model.layers.11.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.315340240485966e-05, "max_abs_err": 0.0301513671875 }, { "name": "model.layers.11.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.9620579627808183e-05, "max_abs_err": 0.12158203125 }, { "name": "model.layers.11.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.986520070815459e-05, "max_abs_err": 0.09228515625 }, { "name": "model.layers.11.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.93460359191522e-05, "max_abs_err": 0.0849609375 }, { "name": "model.layers.11.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.056463512824848e-05, "max_abs_err": 0.142578125 }, { "name": "model.layers.12.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.149574306211434e-05, "max_abs_err": 0.0517578125 }, { "name": "model.layers.12.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.0260714740725234e-05, "max_abs_err": 0.07861328125 }, { "name": "model.layers.12.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.2326930017443374e-05, "max_abs_err": 0.03515625 }, { "name": "model.layers.12.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.9231246773852035e-05, "max_abs_err": 0.12451171875 }, { "name": "model.layers.12.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.7791298129595816e-05, "max_abs_err": 0.10693359375 }, { "name": "model.layers.12.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.152406472712755e-05, "max_abs_err": 0.07080078125 }, { "name": "model.layers.12.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.200970579404384e-05, "max_abs_err": 0.1591796875 }, { "name": "model.layers.13.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.166996950516477e-05, "max_abs_err": 0.058837890625 }, { "name": "model.layers.13.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.0987826650962234e-05, "max_abs_err": 0.10009765625 }, { "name": "model.layers.13.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.9274018238065764e-05, "max_abs_err": 0.035888671875 }, { "name": "model.layers.13.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.74716219084803e-05, "max_abs_err": 0.0673828125 }, { "name": "model.layers.13.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.5125456381356344e-05, "max_abs_err": 0.076171875 }, { "name": "model.layers.13.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.2975319956894964e-05, "max_abs_err": 0.111328125 }, { "name": "model.layers.13.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.4081614760216326e-05, "max_abs_err": 0.15625 }, { "name": "model.layers.14.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.076621917192824e-05, "max_abs_err": 0.052734375 }, { "name": "model.layers.14.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.99992277380079e-05, "max_abs_err": 0.044921875 }, { "name": "model.layers.14.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.784471093444154e-05, "max_abs_err": 0.0284423828125 }, { "name": "model.layers.14.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.642589945229702e-05, "max_abs_err": 0.08203125 }, { "name": "model.layers.14.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.3767449824372306e-05, "max_abs_err": 0.09521484375 }, { "name": "model.layers.14.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.193425047560595e-05, "max_abs_err": 0.068359375 }, { "name": "model.layers.14.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.31473963544704e-05, "max_abs_err": 0.166015625 }, { "name": "model.layers.15.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.199165778118186e-05, "max_abs_err": 0.0556640625 }, { "name": "model.layers.15.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.046514368383214e-05, "max_abs_err": 0.05712890625 }, { "name": "model.layers.15.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.623009226634167e-05, "max_abs_err": 0.0302734375 }, { "name": "model.layers.15.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.4494285500841215e-05, "max_abs_err": 0.061279296875 }, { "name": "model.layers.15.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.115487899980508e-05, "max_abs_err": 0.1015625 }, { "name": "model.layers.15.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.109890141990036e-05, "max_abs_err": 0.076171875 }, { "name": "model.layers.15.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.330993760959245e-05, "max_abs_err": 0.1357421875 }, { "name": "model.layers.16.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.941296017728746e-05, "max_abs_err": 0.054931640625 }, { "name": "model.layers.16.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.736523285624571e-05, "max_abs_err": 0.054931640625 }, { "name": "model.layers.16.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.652102870750241e-05, "max_abs_err": 0.0250244140625 }, { "name": "model.layers.16.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.6339402868179604e-05, "max_abs_err": 0.11279296875 }, { "name": "model.layers.16.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.300264845369384e-05, "max_abs_err": 0.0888671875 }, { "name": "model.layers.16.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.0619535866426304e-05, "max_abs_err": 0.166015625 }, { "name": "model.layers.16.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.1376380977453664e-05, "max_abs_err": 0.203125 }, { "name": "model.layers.17.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.1493825847283006e-05, "max_abs_err": 0.054443359375 }, { "name": "model.layers.17.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.576754534151405e-05, "max_abs_err": 0.043701171875 }, { "name": "model.layers.17.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.444893420441076e-05, "max_abs_err": 0.0302734375 }, { "name": "model.layers.17.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.531174297677353e-05, "max_abs_err": 0.07958984375 }, { "name": "model.layers.17.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.176764196017757e-05, "max_abs_err": 0.1064453125 }, { "name": "model.layers.17.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.049547351314686e-05, "max_abs_err": 0.08154296875 }, { "name": "model.layers.17.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.18923664255999e-05, "max_abs_err": 0.1181640625 }, { "name": "model.layers.18.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.861560617224313e-05, "max_abs_err": 0.054931640625 }, { "name": "model.layers.18.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.588425170164555e-05, "max_abs_err": 0.04296875 }, { "name": "model.layers.18.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.5427233999362215e-05, "max_abs_err": 0.0289306640625 }, { "name": "model.layers.18.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.514937270665541e-05, "max_abs_err": 0.06640625 }, { "name": "model.layers.18.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.116369382245466e-05, "max_abs_err": 0.1083984375 }, { "name": "model.layers.18.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.079098653164692e-05, "max_abs_err": 0.080078125 }, { "name": "model.layers.18.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.228352554491721e-05, "max_abs_err": 0.10986328125 }, { "name": "model.layers.19.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.036000609630719e-05, "max_abs_err": 0.06396484375 }, { "name": "model.layers.19.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.3850661566248164e-05, "max_abs_err": 0.06640625 }, { "name": "model.layers.19.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.384072235552594e-05, "max_abs_err": 0.027587890625 }, { "name": "model.layers.19.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.62527571269311e-05, "max_abs_err": 0.0908203125 }, { "name": "model.layers.19.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.0813600207911804e-05, "max_abs_err": 0.0888671875 }, { "name": "model.layers.19.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.1082679672399536e-05, "max_abs_err": 0.150390625 }, { "name": "model.layers.19.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.245499076205306e-05, "max_abs_err": 0.10888671875 }, { "name": "model.layers.20.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.290465949452482e-05, "max_abs_err": 0.056396484375 }, { "name": "model.layers.20.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.4399468859191984e-05, "max_abs_err": 0.053466796875 }, { "name": "model.layers.20.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.4157532101962715e-05, "max_abs_err": 0.0301513671875 }, { "name": "model.layers.20.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.414571983739734e-05, "max_abs_err": 0.06787109375 }, { "name": "model.layers.20.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.962981645599939e-05, "max_abs_err": 0.08203125 }, { "name": "model.layers.20.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.0833052480593324e-05, "max_abs_err": 0.09814453125 }, { "name": "model.layers.20.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.2301631765440106e-05, "max_abs_err": 0.0703125 }, { "name": "model.layers.21.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.127784995944239e-05, "max_abs_err": 0.05810546875 }, { "name": "model.layers.21.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.346712401253171e-05, "max_abs_err": 0.03955078125 }, { "name": "model.layers.21.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.29715437348932e-05, "max_abs_err": 0.045166015625 }, { "name": "model.layers.21.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.35042059305124e-05, "max_abs_err": 0.087890625 }, { "name": "model.layers.21.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 3.872753950417973e-05, "max_abs_err": 0.10986328125 }, { "name": "model.layers.21.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.050981078762561e-05, "max_abs_err": 0.0634765625 }, { "name": "model.layers.21.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.2720374040072784e-05, "max_abs_err": 0.107421875 }, { "name": "model.layers.22.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.964230927522294e-05, "max_abs_err": 0.051513671875 }, { "name": "model.layers.22.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.223348903702572e-05, "max_abs_err": 0.11572265625 }, { "name": "model.layers.22.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.407497544889338e-05, "max_abs_err": 0.041015625 }, { "name": "model.layers.22.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.4860036976169795e-05, "max_abs_err": 0.078125 }, { "name": "model.layers.22.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.1502094973111525e-05, "max_abs_err": 0.1201171875 }, { "name": "model.layers.22.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.240297130309045e-05, "max_abs_err": 0.080078125 }, { "name": "model.layers.22.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.376686774776317e-05, "max_abs_err": 0.11328125 }, { "name": "model.layers.23.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.5612716348841786e-05, "max_abs_err": 0.06787109375 }, { "name": "model.layers.23.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.4836193663068116e-05, "max_abs_err": 0.04296875 }, { "name": "model.layers.23.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.679287303588353e-05, "max_abs_err": 0.033935546875 }, { "name": "model.layers.23.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.734790880116634e-05, "max_abs_err": 0.056640625 }, { "name": "model.layers.23.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.415847797645256e-05, "max_abs_err": 0.08154296875 }, { "name": "model.layers.23.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.2154188122367486e-05, "max_abs_err": 0.053955078125 }, { "name": "model.layers.23.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.302275556256063e-05, "max_abs_err": 0.0859375 }, { "name": "model.layers.24.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.6409528295043856e-05, "max_abs_err": 0.056884765625 }, { "name": "model.layers.24.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.5104902053717524e-05, "max_abs_err": 0.03955078125 }, { "name": "model.layers.24.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.5672037231270224e-05, "max_abs_err": 0.0419921875 }, { "name": "model.layers.24.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.827424006885849e-05, "max_abs_err": 0.05712890625 }, { "name": "model.layers.24.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.561075911624357e-05, "max_abs_err": 0.080078125 }, { "name": "model.layers.24.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.219940092298202e-05, "max_abs_err": 0.061279296875 }, { "name": "model.layers.24.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.245623495080508e-05, "max_abs_err": 0.103515625 }, { "name": "model.layers.25.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.029790579807013e-05, "max_abs_err": 0.048828125 }, { "name": "model.layers.25.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.5958964872406796e-05, "max_abs_err": 0.0400390625 }, { "name": "model.layers.25.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.674059891840443e-05, "max_abs_err": 0.0322265625 }, { "name": "model.layers.25.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.6622881452785805e-05, "max_abs_err": 0.03173828125 }, { "name": "model.layers.25.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.6418510464718565e-05, "max_abs_err": 0.1298828125 }, { "name": "model.layers.25.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.214714135741815e-05, "max_abs_err": 0.06005859375 }, { "name": "model.layers.25.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.244958472554572e-05, "max_abs_err": 0.07568359375 }, { "name": "model.layers.26.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.228576290188357e-05, "max_abs_err": 0.051025390625 }, { "name": "model.layers.26.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.530592584866099e-05, "max_abs_err": 0.04052734375 }, { "name": "model.layers.26.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.6898981963749975e-05, "max_abs_err": 0.0302734375 }, { "name": "model.layers.26.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.8567755836993456e-05, "max_abs_err": 0.0380859375 }, { "name": "model.layers.26.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.6179287892300636e-05, "max_abs_err": 0.0966796875 }, { "name": "model.layers.26.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.2811276216525584e-05, "max_abs_err": 0.1298828125 }, { "name": "model.layers.26.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.229753903928213e-05, "max_abs_err": 0.07958984375 }, { "name": "model.layers.27.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.07258776249364e-05, "max_abs_err": 0.0615234375 }, { "name": "model.layers.27.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.253531758673489e-05, "max_abs_err": 0.038330078125 }, { "name": "model.layers.27.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.471566353458911e-05, "max_abs_err": 0.03759765625 }, { "name": "model.layers.27.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.705237031681463e-05, "max_abs_err": 0.048583984375 }, { "name": "model.layers.27.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.566908683045767e-05, "max_abs_err": 0.08154296875 }, { "name": "model.layers.27.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.379346501082182e-05, "max_abs_err": 0.072265625 }, { "name": "model.layers.27.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.316532431403175e-05, "max_abs_err": 0.07275390625 }, { "name": "model.layers.28.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.2057108657900244e-05, "max_abs_err": 0.055419921875 }, { "name": "model.layers.28.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.449210635153577e-05, "max_abs_err": 0.037841796875 }, { "name": "model.layers.28.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.798913505510427e-05, "max_abs_err": 0.02783203125 }, { "name": "model.layers.28.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.646681579994038e-05, "max_abs_err": 0.05810546875 }, { "name": "model.layers.28.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.4700973376166075e-05, "max_abs_err": 0.09375 }, { "name": "model.layers.28.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.467098551685922e-05, "max_abs_err": 0.109375 }, { "name": "model.layers.28.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.3978128815069795e-05, "max_abs_err": 0.123046875 }, { "name": "model.layers.29.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.713815724244341e-05, "max_abs_err": 0.059326171875 }, { "name": "model.layers.29.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.411524448893033e-05, "max_abs_err": 0.048828125 }, { "name": "model.layers.29.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.480330244405195e-05, "max_abs_err": 0.052978515625 }, { "name": "model.layers.29.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.9386835851473734e-05, "max_abs_err": 0.037109375 }, { "name": "model.layers.29.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.4326778152026236e-05, "max_abs_err": 0.11279296875 }, { "name": "model.layers.29.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.587063813232817e-05, "max_abs_err": 0.064453125 }, { "name": "model.layers.29.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.4456948671722785e-05, "max_abs_err": 0.07568359375 }, { "name": "model.layers.30.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.963075505453162e-05, "max_abs_err": 0.0751953125 }, { "name": "model.layers.30.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.592043503886089e-05, "max_abs_err": 0.0478515625 }, { "name": "model.layers.30.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 4.95524363941513e-05, "max_abs_err": 0.04248046875 }, { "name": "model.layers.30.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.8655944081256166e-05, "max_abs_err": 0.041015625 }, { "name": "model.layers.30.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.3836113036377355e-05, "max_abs_err": 0.099609375 }, { "name": "model.layers.30.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.6930828830227256e-05, "max_abs_err": 0.064453125 }, { "name": "model.layers.30.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.567719224723987e-05, "max_abs_err": 0.1015625 }, { "name": "model.layers.31.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.13320267398376e-05, "max_abs_err": 0.06689453125 }, { "name": "model.layers.31.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.378564360900782e-05, "max_abs_err": 0.042236328125 }, { "name": "model.layers.31.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.362384399631992e-05, "max_abs_err": 0.03466796875 }, { "name": "model.layers.31.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 4.123463077121414e-05, "max_abs_err": 0.0546875 }, { "name": "model.layers.31.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.295111648389138e-05, "max_abs_err": 0.10400390625 }, { "name": "model.layers.31.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.8131256335182115e-05, "max_abs_err": 0.09326171875 }, { "name": "model.layers.31.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.738035568152554e-05, "max_abs_err": 0.12109375 }, { "name": "model.layers.32.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.0278038795804605e-05, "max_abs_err": 0.0703125 }, { "name": "model.layers.32.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.3849697501864284e-05, "max_abs_err": 0.04052734375 }, { "name": "model.layers.32.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.385827898862772e-05, "max_abs_err": 0.03857421875 }, { "name": "model.layers.32.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 4.109629662707448e-05, "max_abs_err": 0.0712890625 }, { "name": "model.layers.32.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.217680907458998e-05, "max_abs_err": 0.11572265625 }, { "name": "model.layers.32.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.854234794038348e-05, "max_abs_err": 0.060546875 }, { "name": "model.layers.32.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.764810000779107e-05, "max_abs_err": 0.11279296875 }, { "name": "model.layers.33.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 4.1338440496474504e-05, "max_abs_err": 0.06298828125 }, { "name": "model.layers.33.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 2.7914715246879496e-05, "max_abs_err": 0.034423828125 }, { "name": "model.layers.33.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 6.060989107936621e-05, "max_abs_err": 0.039306640625 }, { "name": "model.layers.33.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 4.454971713130362e-05, "max_abs_err": 0.111328125 }, { "name": "model.layers.33.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.177831578999758e-05, "max_abs_err": 0.1025390625 }, { "name": "model.layers.33.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.9026202759705484e-05, "max_abs_err": 0.072265625 }, { "name": "model.layers.33.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.6996527089504525e-05, "max_abs_err": 0.0986328125 }, { "name": "model.layers.34.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.781018313020468e-05, "max_abs_err": 0.06884765625 }, { "name": "model.layers.34.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.1367559131467715e-05, "max_abs_err": 0.056640625 }, { "name": "model.layers.34.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.993696322548203e-05, "max_abs_err": 0.04052734375 }, { "name": "model.layers.34.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 4.2564533941913396e-05, "max_abs_err": 0.0595703125 }, { "name": "model.layers.34.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.3353076762286946e-05, "max_abs_err": 0.1015625 }, { "name": "model.layers.34.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.783711119671352e-05, "max_abs_err": 0.0869140625 }, { "name": "model.layers.34.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.394195639179088e-05, "max_abs_err": 0.125 }, { "name": "model.layers.35.self_attn.q_proj", "shape": [ 4096, 2560 ], "dtype": "bfloat16", "mse": 3.766561712836847e-05, "max_abs_err": 0.052001953125 }, { "name": "model.layers.35.self_attn.k_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 3.516471042530611e-05, "max_abs_err": 0.055908203125 }, { "name": "model.layers.35.self_attn.v_proj", "shape": [ 1024, 2560 ], "dtype": "bfloat16", "mse": 5.2958821470383555e-05, "max_abs_err": 0.034423828125 }, { "name": "model.layers.35.self_attn.o_proj", "shape": [ 2560, 4096 ], "dtype": "bfloat16", "mse": 3.99371529056225e-05, "max_abs_err": 0.0869140625 }, { "name": "model.layers.35.mlp.gate_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.7669887862866744e-05, "max_abs_err": 0.134765625 }, { "name": "model.layers.35.mlp.up_proj", "shape": [ 9728, 2560 ], "dtype": "bfloat16", "mse": 4.911969153909013e-05, "max_abs_err": 0.1337890625 }, { "name": "model.layers.35.mlp.down_proj", "shape": [ 2560, 9728 ], "dtype": "bfloat16", "mse": 4.3243362597422674e-05, "max_abs_err": 0.1298828125 } ], "source_model": "Qwen/Qwen3-4B-Thinking-2507", "dtype": "bf16", "target_regex": ".*", "ignore_regex": [], "include_lm_head": false, "note": "Dense fake quant: rounded weights are saved as normal HF tensors." }