Valid precision + execution provider combinations are: FP32 CPU, FP32 CUDA, FP16 CUDA, FP16 DML, FP16 TRT-RTX, BF16 CUDA, BF16 TRT-RTX, INT8 CPU, INT8 CUDA, INT8 WebGPU, INT4 CPU, INT4 CUDA, INT4 DML, INT4 WebGPU Extra options: {} GroupQueryAttention (GQA) is used in this model. GroupQueryAttention (GQA) is used in this model. Loading weights: 0%| | 0/426 [00:00 mtp.onnx Saving ONNX model in /home/anon/knowline/onnx/model 0it [00:00, ?it/s] Saving model.layers.0.input_layernorm.weight (f16, [2560]): 0%| | 1/404 [00:00<00:01, 257.64it/s] Saving model.layers.0.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 0%| | 2/404 [00:00<00:00, 479.76it/s] Saving model.layers.0.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 1%| | 3/404 [00:00<00:03, 102.33it/s] Saving model.layers.0.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 1%| | 4/404 [00:00<00:04, 92.50it/s] Saving model.layers.0.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 1%| | 5/404 [00:00<00:03, 114.79it/s] Saving model.layers.0.linear_attn.conv1d.weight (f16, [8192,1,4]): 1%|▏ | 6/404 [00:00<00:02, 137.16it/s] Saving model.layers.0.linear_attn.conv1d.bias (f16, [8192]): 2%|▏ | 7/404 [00:00<00:02, 159.58it/s] Saving model.layers.0.linear_attn.A_log (f32, [32]): 2%|▏ | 8/404 [00:00<00:02, 181.96it/s] Saving model.layers.0.linear_attn.dt_bias (f32, [32]): 2%|▏ | 9/404 [00:00<00:01, 204.30it/s] Saving model.layers.0.linear_attn.norm.weight (f16, [128]): 2%|▏ | 10/404 [00:00<00:01, 226.72it/s] Saving model.layers.0.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 3%|▎ | 11/404 [00:00<00:01, 249.07it/s] Saving model.layers.0.post_attention_layernorm.weight (f16, [2560]): 3%|▎ | 12/404 [00:00<00:01, 208.07it/s] Saving model.layers.0.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 3%|▎ | 13/404 [00:00<00:01, 224.36it/s] Saving model.layers.0.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 3%|▎ | 14/404 [00:00<00:13, 29.58it/s] Saving model.layers.0.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 3%|▎ | 14/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.input_layernorm.weight (f16, [2560]): 4%|▎ | 15/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 4%|▍ | 16/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 4%|▍ | 17/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 4%|▍ | 18/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 5%|▍ | 19/404 [00:00<00:13, 29.58it/s] Saving model.layers.1.linear_attn.conv1d.weight (f16, [8192,1,4]): 5%|▍ | 20/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.linear_attn.conv1d.bias (f16, [8192]): 5%|▌ | 21/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.linear_attn.A_log (f32, [32]): 5%|▌ | 22/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.linear_attn.dt_bias (f32, [32]): 6%|▌ | 23/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.linear_attn.norm.weight (f16, [128]): 6%|▌ | 24/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 6%|▌ | 25/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.post_attention_layernorm.weight (f16, [2560]): 6%|▋ | 26/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 7%|▋ | 27/404 [00:00<00:12, 29.58it/s] Saving model.layers.1.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 7%|▋ | 28/404 [00:00<00:13, 28.42it/s] Saving model.layers.1.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 7%|▋ | 28/404 [00:00<00:13, 28.42it/s] Saving model.layers.2.input_layernorm.weight (f16, [2560]): 7%|▋ | 29/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 7%|▋ | 30/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 8%|▊ | 31/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 8%|▊ | 32/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 8%|▊ | 33/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.conv1d.weight (f16, [8192,1,4]): 8%|▊ | 34/404 [00:01<00:13, 28.42it/s] Saving model.layers.2.linear_attn.conv1d.bias (f16, [8192]): 9%|▊ | 35/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.linear_attn.A_log (f32, [32]): 9%|▉ | 36/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.linear_attn.dt_bias (f32, [32]): 9%|▉ | 37/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.linear_attn.norm.weight (f16, [128]): 9%|▉ | 38/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 10%|▉ | 39/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.post_attention_layernorm.weight (f16, [2560]): 10%|▉ | 40/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 10%|█ | 41/404 [00:01<00:12, 28.42it/s] Saving model.layers.2.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 10%|█ | 42/404 [00:01<00:12, 28.04it/s] Saving model.layers.2.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 10%|█ | 42/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.input_layernorm.weight (f16, [2560]): 11%|█ | 43/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 11%|█ | 44/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.attn.q_norm.layernorm.weight (f16, [256]): 11%|█ | 45/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.attn.k_norm.layernorm.weight (f16, [256]): 11%|█▏ | 46/404 [00:01<00:12, 28.04it/s] Saving cos_cache (f32, [262144,32]): 12%|█▏ | 47/404 [00:01<00:12, 28.04it/s] Saving sin_cache (f32, [262144,32]): 12%|█▏ | 48/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.attn.o_proj.MatMul.weight (f16, [4096,2560]): 12%|█▏ | 49/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.post_attention_layernorm.weight (f16, [2560]): 12%|█▏ | 50/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 13%|█▎ | 51/404 [00:01<00:12, 28.04it/s] Saving model.layers.3.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 13%|█▎ | 52/404 [00:01<00:14, 24.68it/s] Saving model.layers.3.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 13%|█▎ | 52/404 [00:01<00:14, 24.68it/s] Saving model.layers.4.input_layernorm.weight (f16, [2560]): 13%|█▎ | 53/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 13%|█▎ | 54/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 14%|█▎ | 55/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 14%|█▍ | 56/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 14%|█▍ | 57/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.conv1d.weight (f16, [8192,1,4]): 14%|█▍ | 58/404 [00:02<00:14, 24.68it/s] Saving model.layers.4.linear_attn.conv1d.bias (f16, [8192]): 15%|█▍ | 59/404 [00:02<00:13, 24.68it/s] Saving model.layers.4.linear_attn.A_log (f32, [32]): 15%|█▍ | 60/404 [00:02<00:13, 24.68it/s] Saving model.layers.4.linear_attn.dt_bias (f32, [32]): 15%|█▌ | 61/404 [00:02<00:13, 24.68it/s] Saving model.layers.4.linear_attn.norm.weight (f16, [128]): 15%|█▌ | 62/404 [00:02<00:13, 24.68it/s] Saving model.layers.4.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 16%|█▌ | 63/404 [00:02<00:13, 24.68it/s] Saving model.layers.4.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 16%|█▌ | 64/404 [00:02<00:09, 34.15it/s] Saving model.layers.4.post_attention_layernorm.weight (f16, [2560]): 16%|█▌ | 64/404 [00:02<00:09, 34.15it/s] Saving model.layers.4.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 16%|█▌ | 65/404 [00:02<00:09, 34.15it/s] Saving model.layers.4.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 16%|█▋ | 66/404 [00:02<00:09, 34.15it/s] Saving model.layers.5.input_layernorm.weight (f16, [2560]): 17%|█▋ | 67/404 [00:02<00:09, 34.15it/s] Saving model.layers.5.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 17%|█▋ | 68/404 [00:02<00:09, 34.15it/s] Saving model.layers.5.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 17%|█▋ | 69/404 [00:02<00:09, 34.15it/s] Saving model.layers.5.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 17%|█▋ | 70/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 17%|█▋ | 70/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 18%|█▊ | 71/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.conv1d.weight (f16, [8192,1,4]): 18%|█▊ | 72/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.conv1d.bias (f16, [8192]): 18%|█▊ | 73/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.A_log (f32, [32]): 18%|█▊ | 74/404 [00:02<00:13, 25.32it/s] Saving model.layers.5.linear_attn.dt_bias (f32, [32]): 19%|█▊ | 75/404 [00:02<00:12, 25.32it/s] Saving model.layers.5.linear_attn.norm.weight (f16, [128]): 19%|█▉ | 76/404 [00:02<00:12, 25.32it/s] Saving model.layers.5.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 19%|█▉ | 77/404 [00:02<00:12, 25.32it/s] Saving model.layers.5.post_attention_layernorm.weight (f16, [2560]): 19%|█▉ | 78/404 [00:02<00:12, 25.32it/s] Saving model.layers.5.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 20%|█▉ | 79/404 [00:02<00:12, 25.32it/s] Saving model.layers.5.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 20%|█▉ | 80/404 [00:03<00:13, 24.43it/s] Saving model.layers.5.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 20%|█▉ | 80/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.input_layernorm.weight (f16, [2560]): 20%|██ | 81/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 20%|██ | 82/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 21%|██ | 83/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 21%|██ | 84/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 21%|██ | 85/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.conv1d.weight (f16, [8192,1,4]): 21%|██▏ | 86/404 [00:03<00:13, 24.43it/s] Saving model.layers.6.linear_attn.conv1d.bias (f16, [8192]): 22%|██▏ | 87/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.linear_attn.A_log (f32, [32]): 22%|██▏ | 88/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.linear_attn.dt_bias (f32, [32]): 22%|██▏ | 89/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.linear_attn.norm.weight (f16, [128]): 22%|██▏ | 90/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 23%|██▎ | 91/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.post_attention_layernorm.weight (f16, [2560]): 23%|██▎ | 92/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 23%|██▎ | 93/404 [00:03<00:12, 24.43it/s] Saving model.layers.6.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 23%|██▎ | 94/404 [00:03<00:12, 25.59it/s] Saving model.layers.6.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 23%|██▎ | 94/404 [00:03<00:12, 25.59it/s] Saving model.layers.7.input_layernorm.weight (f16, [2560]): 24%|██▎ | 95/404 [00:03<00:12, 25.59it/s] Saving model.layers.7.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 24%|██▍ | 96/404 [00:03<00:12, 25.59it/s] Saving model.layers.7.attn.q_norm.layernorm.weight (f16, [256]): 24%|██▍ | 97/404 [00:03<00:11, 25.59it/s] Saving model.layers.7.attn.k_norm.layernorm.weight (f16, [256]): 24%|██▍ | 98/404 [00:03<00:11, 25.59it/s] Saving model.layers.7.attn.o_proj.MatMul.weight (f16, [4096,2560]): 25%|██▍ | 99/404 [00:03<00:11, 25.59it/s] Saving model.layers.7.post_attention_layernorm.weight (f16, [2560]): 25%|██▍ | 100/404 [00:03<00:11, 25.59it/s] Saving model.layers.7.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 25%|██▌ | 101/404 [00:03<00:11, 25.59it/s] Saving model.layers.7.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 25%|██▌ | 102/404 [00:04<00:13, 22.30it/s] Saving model.layers.7.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 25%|██▌ | 102/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.input_layernorm.weight (f16, [2560]): 25%|██▌ | 103/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 26%|██▌ | 104/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 26%|██▌ | 105/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 26%|██▌ | 106/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 26%|██▋ | 107/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.conv1d.weight (f16, [8192,1,4]): 27%|██▋ | 108/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.conv1d.bias (f16, [8192]): 27%|██▋ | 109/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.A_log (f32, [32]): 27%|██▋ | 110/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.dt_bias (f32, [32]): 27%|██▋ | 111/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.norm.weight (f16, [128]): 28%|██▊ | 112/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 28%|██▊ | 113/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.post_attention_layernorm.weight (f16, [2560]): 28%|██▊ | 114/404 [00:04<00:13, 22.30it/s] Saving model.layers.8.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 28%|██▊ | 115/404 [00:04<00:12, 22.30it/s] Saving model.layers.8.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 29%|██▊ | 116/404 [00:04<00:11, 24.11it/s] Saving model.layers.8.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 29%|██▊ | 116/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.input_layernorm.weight (f16, [2560]): 29%|██▉ | 117/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 29%|██▉ | 118/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 29%|██▉ | 119/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 30%|██▉ | 120/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 30%|██▉ | 121/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.conv1d.weight (f16, [8192,1,4]): 30%|███ | 122/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.conv1d.bias (f16, [8192]): 30%|███ | 123/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.A_log (f32, [32]): 31%|███ | 124/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.dt_bias (f32, [32]): 31%|███ | 125/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.norm.weight (f16, [128]): 31%|███ | 126/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 31%|███▏ | 127/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.post_attention_layernorm.weight (f16, [2560]): 32%|███▏ | 128/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 32%|███▏ | 129/404 [00:04<00:11, 24.11it/s] Saving model.layers.9.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 32%|███▏ | 130/404 [00:05<00:10, 25.22it/s] Saving model.layers.9.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 32%|███▏ | 130/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.input_layernorm.weight (f16, [2560]): 32%|███▏ | 131/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 33%|███▎ | 132/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 33%|███▎ | 133/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 33%|███▎ | 134/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 33%|███▎ | 135/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.conv1d.weight (f16, [8192,1,4]): 34%|███▎ | 136/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.conv1d.bias (f16, [8192]): 34%|███▍ | 137/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.A_log (f32, [32]): 34%|███▍ | 138/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.dt_bias (f32, [32]): 34%|███▍ | 139/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.norm.weight (f16, [128]): 35%|███▍ | 140/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 35%|███▍ | 141/404 [00:05<00:10, 25.22it/s] Saving model.layers.10.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 35%|███▌ | 142/404 [00:05<00:07, 33.03it/s] Saving model.layers.10.post_attention_layernorm.weight (f16, [2560]): 35%|███▌ | 142/404 [00:05<00:07, 33.03it/s] Saving model.layers.10.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 35%|███▌ | 143/404 [00:05<00:07, 33.03it/s] Saving model.layers.10.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 36%|███▌ | 144/404 [00:05<00:07, 33.03it/s] Saving model.layers.11.input_layernorm.weight (f16, [2560]): 36%|███▌ | 145/404 [00:05<00:07, 33.03it/s] Saving model.layers.11.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 36%|███▌ | 146/404 [00:05<00:07, 33.03it/s] Saving model.layers.11.attn.q_norm.layernorm.weight (f16, [256]): 36%|███▋ | 147/404 [00:05<00:07, 33.03it/s] Saving model.layers.11.attn.q_norm.layernorm.weight (f16, [256]): 37%|███▋ | 148/404 [00:05<00:09, 26.06it/s] Saving model.layers.11.attn.k_norm.layernorm.weight (f16, [256]): 37%|███▋ | 148/404 [00:05<00:09, 26.06it/s] Saving model.layers.11.attn.o_proj.MatMul.weight (f16, [4096,2560]): 37%|███▋ | 149/404 [00:05<00:09, 26.06it/s] Saving model.layers.11.post_attention_layernorm.weight (f16, [2560]): 37%|███▋ | 150/404 [00:05<00:09, 26.06it/s] Saving model.layers.11.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 37%|███▋ | 151/404 [00:05<00:09, 26.06it/s] Saving model.layers.11.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 38%|███▊ | 152/404 [00:06<00:09, 26.06it/s] Saving model.layers.11.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 38%|███▊ | 153/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.input_layernorm.weight (f16, [2560]): 38%|███▊ | 153/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 38%|███▊ | 154/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 38%|███▊ | 155/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 39%|███▊ | 156/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 39%|███▉ | 157/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.conv1d.weight (f16, [8192,1,4]): 39%|███▉ | 158/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.conv1d.bias (f16, [8192]): 39%|███▉ | 159/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.A_log (f32, [32]): 40%|███▉ | 160/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.dt_bias (f32, [32]): 40%|███▉ | 161/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.norm.weight (f16, [128]): 40%|████ | 162/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 40%|████ | 163/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.post_attention_layernorm.weight (f16, [2560]): 41%|████ | 164/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 41%|████ | 165/404 [00:06<00:11, 21.15it/s] Saving model.layers.12.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 41%|████ | 166/404 [00:06<00:10, 23.45it/s] Saving model.layers.12.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 41%|████ | 166/404 [00:06<00:10, 23.45it/s] Saving model.layers.13.input_layernorm.weight (f16, [2560]): 41%|████▏ | 167/404 [00:06<00:10, 23.45it/s] Saving model.layers.13.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 42%|████▏ | 168/404 [00:06<00:10, 23.45it/s] Saving model.layers.13.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 42%|████▏ | 169/404 [00:06<00:10, 23.45it/s] Saving model.layers.13.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 42%|████▏ | 170/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 42%|████▏ | 171/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.conv1d.weight (f16, [8192,1,4]): 43%|████▎ | 172/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.conv1d.bias (f16, [8192]): 43%|████▎ | 173/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.A_log (f32, [32]): 43%|████▎ | 174/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.dt_bias (f32, [32]): 43%|████▎ | 175/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.norm.weight (f16, [128]): 44%|████▎ | 176/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 44%|████▍ | 177/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.post_attention_layernorm.weight (f16, [2560]): 44%|████▍ | 178/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 44%|████▍ | 179/404 [00:06<00:09, 23.45it/s] Saving model.layers.13.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 45%|████▍ | 180/404 [00:06<00:08, 25.09it/s] Saving model.layers.13.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 45%|████▍ | 180/404 [00:06<00:08, 25.09it/s] Saving model.layers.14.input_layernorm.weight (f16, [2560]): 45%|████▍ | 181/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 45%|████▌ | 182/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 45%|████▌ | 183/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 46%|████▌ | 184/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 46%|████▌ | 185/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.conv1d.weight (f16, [8192,1,4]): 46%|████▌ | 186/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.conv1d.bias (f16, [8192]): 46%|████▋ | 187/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.A_log (f32, [32]): 47%|████▋ | 188/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.dt_bias (f32, [32]): 47%|████▋ | 189/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.norm.weight (f16, [128]): 47%|████▋ | 190/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 47%|████▋ | 191/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.post_attention_layernorm.weight (f16, [2560]): 48%|████▊ | 192/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 48%|████▊ | 193/404 [00:07<00:08, 25.09it/s] Saving model.layers.14.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 48%|████▊ | 194/404 [00:07<00:08, 25.86it/s] Saving model.layers.14.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 48%|████▊ | 194/404 [00:07<00:08, 25.86it/s] Saving model.layers.15.input_layernorm.weight (f16, [2560]): 48%|████▊ | 195/404 [00:07<00:08, 25.86it/s] Saving model.layers.15.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 49%|████▊ | 196/404 [00:07<00:08, 25.86it/s] Saving model.layers.15.attn.q_norm.layernorm.weight (f16, [256]): 49%|████▉ | 197/404 [00:07<00:08, 25.86it/s] Saving model.layers.15.attn.k_norm.layernorm.weight (f16, [256]): 49%|████▉ | 198/404 [00:07<00:07, 25.86it/s] Saving model.layers.15.attn.o_proj.MatMul.weight (f16, [4096,2560]): 49%|████▉ | 199/404 [00:07<00:07, 25.86it/s] Saving model.layers.15.post_attention_layernorm.weight (f16, [2560]): 50%|████▉ | 200/404 [00:07<00:07, 25.86it/s] Saving model.layers.15.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 50%|████▉ | 201/404 [00:07<00:07, 25.86it/s] Saving model.layers.15.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 50%|█████ | 202/404 [00:07<00:08, 23.13it/s] Saving model.layers.15.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 50%|█████ | 202/404 [00:07<00:08, 23.13it/s] Saving model.layers.16.input_layernorm.weight (f16, [2560]): 50%|█████ | 203/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 50%|█████ | 204/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 51%|█████ | 205/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 51%|█████ | 206/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 51%|█████ | 207/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.conv1d.weight (f16, [8192,1,4]): 51%|█████▏ | 208/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.conv1d.bias (f16, [8192]): 52%|█████▏ | 209/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.A_log (f32, [32]): 52%|█████▏ | 210/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.dt_bias (f32, [32]): 52%|█████▏ | 211/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.norm.weight (f16, [128]): 52%|█████▏ | 212/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 53%|█████▎ | 213/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.post_attention_layernorm.weight (f16, [2560]): 53%|█████▎ | 214/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 53%|█████▎ | 215/404 [00:08<00:08, 23.13it/s] Saving model.layers.16.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 53%|█████▎ | 216/404 [00:08<00:07, 24.91it/s] Saving model.layers.16.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 53%|█████▎ | 216/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.input_layernorm.weight (f16, [2560]): 54%|█████▎ | 217/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 54%|█████▍ | 218/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 54%|█████▍ | 219/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 54%|█████▍ | 220/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 55%|█████▍ | 221/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.conv1d.weight (f16, [8192,1,4]): 55%|█████▍ | 222/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.conv1d.bias (f16, [8192]): 55%|█████▌ | 223/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.A_log (f32, [32]): 55%|█████▌ | 224/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.dt_bias (f32, [32]): 56%|█████▌ | 225/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.norm.weight (f16, [128]): 56%|█████▌ | 226/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 56%|█████▌ | 227/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.post_attention_layernorm.weight (f16, [2560]): 56%|█████▋ | 228/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 57%|█████▋ | 229/404 [00:08<00:07, 24.91it/s] Saving model.layers.17.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 57%|█████▋ | 230/404 [00:08<00:06, 26.16it/s] Saving model.layers.17.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 57%|█████▋ | 230/404 [00:08<00:06, 26.16it/s] Saving model.layers.18.input_layernorm.weight (f16, [2560]): 57%|█████▋ | 231/404 [00:08<00:06, 26.16it/s] Saving model.layers.18.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 57%|█████▋ | 232/404 [00:08<00:06, 26.16it/s] Saving model.layers.18.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 58%|█████▊ | 233/404 [00:08<00:06, 26.16it/s] Saving model.layers.18.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 58%|█████▊ | 234/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 58%|█████▊ | 235/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.conv1d.weight (f16, [8192,1,4]): 58%|█████▊ | 236/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.conv1d.bias (f16, [8192]): 59%|█████▊ | 237/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.A_log (f32, [32]): 59%|█████▉ | 238/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.dt_bias (f32, [32]): 59%|█████▉ | 239/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.norm.weight (f16, [128]): 59%|█████▉ | 240/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 60%|█████▉ | 241/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.post_attention_layernorm.weight (f16, [2560]): 60%|█████▉ | 242/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 60%|██████ | 243/404 [00:09<00:06, 26.16it/s] Saving model.layers.18.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 60%|██████ | 244/404 [00:09<00:05, 27.04it/s] Saving model.layers.18.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 60%|██████ | 244/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.input_layernorm.weight (f16, [2560]): 61%|██████ | 245/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 61%|██████ | 246/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.attn.q_norm.layernorm.weight (f16, [256]): 61%|██████ | 247/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.attn.k_norm.layernorm.weight (f16, [256]): 61%|██████▏ | 248/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.attn.o_proj.MatMul.weight (f16, [4096,2560]): 62%|██████▏ | 249/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.post_attention_layernorm.weight (f16, [2560]): 62%|██████▏ | 250/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 62%|██████▏ | 251/404 [00:09<00:05, 27.04it/s] Saving model.layers.19.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 62%|██████▏ | 252/404 [00:09<00:06, 24.13it/s] Saving model.layers.19.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 62%|██████▏ | 252/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.input_layernorm.weight (f16, [2560]): 63%|██████▎ | 253/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 63%|██████▎ | 254/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 63%|██████▎ | 255/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 63%|██████▎ | 256/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 64%|██████▎ | 257/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.conv1d.weight (f16, [8192,1,4]): 64%|██████▍ | 258/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.conv1d.bias (f16, [8192]): 64%|██████▍ | 259/404 [00:09<00:06, 24.13it/s] Saving model.layers.20.linear_attn.A_log (f32, [32]): 64%|██████▍ | 260/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.linear_attn.dt_bias (f32, [32]): 65%|██████▍ | 261/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.linear_attn.norm.weight (f16, [128]): 65%|██████▍ | 262/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 65%|██████▌ | 263/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.post_attention_layernorm.weight (f16, [2560]): 65%|██████▌ | 264/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 66%|██████▌ | 265/404 [00:09<00:05, 24.13it/s] Saving model.layers.20.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 66%|██████▌ | 266/404 [00:10<00:05, 25.38it/s] Saving model.layers.20.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 66%|██████▌ | 266/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.input_layernorm.weight (f16, [2560]): 66%|██████▌ | 267/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 66%|██████▋ | 268/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 67%|██████▋ | 269/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 67%|██████▋ | 270/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 67%|██████▋ | 271/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.conv1d.weight (f16, [8192,1,4]): 67%|██████▋ | 272/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.conv1d.bias (f16, [8192]): 68%|██████▊ | 273/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.A_log (f32, [32]): 68%|██████▊ | 274/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.dt_bias (f32, [32]): 68%|██████▊ | 275/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.norm.weight (f16, [128]): 68%|██████▊ | 276/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 69%|██████▊ | 277/404 [00:10<00:05, 25.38it/s] Saving model.layers.21.post_attention_layernorm.weight (f16, [2560]): 69%|██████▉ | 278/404 [00:10<00:04, 25.38it/s] Saving model.layers.21.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 69%|██████▉ | 279/404 [00:10<00:04, 25.38it/s] Saving model.layers.21.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 69%|██████▉ | 280/404 [00:10<00:04, 26.50it/s] Saving model.layers.21.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 69%|██████▉ | 280/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.input_layernorm.weight (f16, [2560]): 70%|██████▉ | 281/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 70%|██████▉ | 282/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 70%|███████ | 283/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 70%|███████ | 284/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 71%|███████ | 285/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.conv1d.weight (f16, [8192,1,4]): 71%|███████ | 286/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.conv1d.bias (f16, [8192]): 71%|███████ | 287/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.A_log (f32, [32]): 71%|███████▏ | 288/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.dt_bias (f32, [32]): 72%|███████▏ | 289/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.norm.weight (f16, [128]): 72%|███████▏ | 290/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 72%|███████▏ | 291/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.post_attention_layernorm.weight (f16, [2560]): 72%|███████▏ | 292/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 73%|███████▎ | 293/404 [00:10<00:04, 26.50it/s] Saving model.layers.22.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 73%|███████▎ | 294/404 [00:11<00:03, 27.68it/s] Saving model.layers.22.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 73%|███████▎ | 294/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.input_layernorm.weight (f16, [2560]): 73%|███████▎ | 295/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 73%|███████▎ | 296/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.attn.q_norm.layernorm.weight (f16, [256]): 74%|███████▎ | 297/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.attn.k_norm.layernorm.weight (f16, [256]): 74%|███████▍ | 298/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.attn.o_proj.MatMul.weight (f16, [4096,2560]): 74%|███████▍ | 299/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.post_attention_layernorm.weight (f16, [2560]): 74%|███████▍ | 300/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 75%|███████▍ | 301/404 [00:11<00:03, 27.68it/s] Saving model.layers.23.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 75%|███████▍ | 302/404 [00:11<00:04, 24.59it/s] Saving model.layers.23.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 75%|███████▍ | 302/404 [00:11<00:04, 24.59it/s] Saving model.layers.24.input_layernorm.weight (f16, [2560]): 75%|███████▌ | 303/404 [00:11<00:04, 24.59it/s] Saving model.layers.24.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 75%|███████▌ | 304/404 [00:11<00:04, 24.59it/s] Saving model.layers.24.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 75%|███████▌ | 305/404 [00:11<00:04, 24.59it/s] Saving model.layers.24.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 76%|███████▌ | 306/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 76%|███████▌ | 307/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.conv1d.weight (f16, [8192,1,4]): 76%|███████▌ | 308/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.conv1d.bias (f16, [8192]): 76%|███████▋ | 309/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.A_log (f32, [32]): 77%|███████▋ | 310/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.dt_bias (f32, [32]): 77%|███████▋ | 311/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.norm.weight (f16, [128]): 77%|███████▋ | 312/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 77%|███████▋ | 313/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.post_attention_layernorm.weight (f16, [2560]): 78%|███████▊ | 314/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 78%|███████▊ | 315/404 [00:11<00:03, 24.59it/s] Saving model.layers.24.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 78%|███████▊ | 316/404 [00:12<00:03, 25.91it/s] Saving model.layers.24.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 78%|███████▊ | 316/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.input_layernorm.weight (f16, [2560]): 78%|███████▊ | 317/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 79%|███████▊ | 318/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 79%|███████▉ | 319/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 79%|███████▉ | 320/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 79%|███████▉ | 321/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.conv1d.weight (f16, [8192,1,4]): 80%|███████▉ | 322/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.conv1d.bias (f16, [8192]): 80%|███████▉ | 323/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.A_log (f32, [32]): 80%|████████ | 324/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.dt_bias (f32, [32]): 80%|████████ | 325/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.norm.weight (f16, [128]): 81%|████████ | 326/404 [00:12<00:03, 25.91it/s] Saving model.layers.25.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 81%|████████ | 327/404 [00:12<00:02, 25.91it/s] Saving model.layers.25.post_attention_layernorm.weight (f16, [2560]): 81%|████████ | 328/404 [00:12<00:02, 25.91it/s] Saving model.layers.25.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 81%|████████▏ | 329/404 [00:12<00:02, 25.91it/s] Saving model.layers.25.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 82%|████████▏ | 330/404 [00:12<00:02, 26.86it/s] Saving model.layers.25.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 82%|████████▏ | 330/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.input_layernorm.weight (f16, [2560]): 82%|████████▏ | 331/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 82%|████████▏ | 332/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 82%|████████▏ | 333/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 83%|████████▎ | 334/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 83%|████████▎ | 335/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.conv1d.weight (f16, [8192,1,4]): 83%|████████▎ | 336/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.conv1d.bias (f16, [8192]): 83%|████████▎ | 337/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.A_log (f32, [32]): 84%|████████▎ | 338/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.dt_bias (f32, [32]): 84%|████████▍ | 339/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.norm.weight (f16, [128]): 84%|████████▍ | 340/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 84%|████████▍ | 341/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.post_attention_layernorm.weight (f16, [2560]): 85%|████████▍ | 342/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 85%|████████▍ | 343/404 [00:12<00:02, 26.86it/s] Saving model.layers.26.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 85%|████████▌ | 344/404 [00:13<00:02, 27.08it/s] Saving model.layers.26.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 85%|████████▌ | 344/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.input_layernorm.weight (f16, [2560]): 85%|████████▌ | 345/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 86%|████████▌ | 346/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.attn.q_norm.layernorm.weight (f16, [256]): 86%|████████▌ | 347/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.attn.k_norm.layernorm.weight (f16, [256]): 86%|████████▌ | 348/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.attn.o_proj.MatMul.weight (f16, [4096,2560]): 86%|████████▋ | 349/404 [00:13<00:02, 27.08it/s] Saving model.layers.27.post_attention_layernorm.weight (f16, [2560]): 87%|████████▋ | 350/404 [00:13<00:01, 27.08it/s] Saving model.layers.27.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 87%|████████▋ | 351/404 [00:13<00:01, 27.08it/s] Saving model.layers.27.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 87%|████████▋ | 352/404 [00:13<00:02, 23.82it/s] Saving model.layers.27.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 87%|████████▋ | 352/404 [00:13<00:02, 23.82it/s] Saving model.layers.28.input_layernorm.weight (f16, [2560]): 87%|████████▋ | 353/404 [00:13<00:02, 23.82it/s] Saving model.layers.28.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 88%|████████▊ | 354/404 [00:13<00:02, 23.82it/s] Saving model.layers.28.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 88%|████████▊ | 355/404 [00:13<00:02, 23.82it/s] Saving model.layers.28.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 88%|████████▊ | 356/404 [00:13<00:02, 23.82it/s] Saving model.layers.28.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 88%|████████▊ | 357/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.conv1d.weight (f16, [8192,1,4]): 89%|████████▊ | 358/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.conv1d.bias (f16, [8192]): 89%|████████▉ | 359/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.A_log (f32, [32]): 89%|████████▉ | 360/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.dt_bias (f32, [32]): 89%|████████▉ | 361/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.norm.weight (f16, [128]): 90%|████████▉ | 362/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 90%|████████▉ | 363/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.post_attention_layernorm.weight (f16, [2560]): 90%|█████████ | 364/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 90%|█████████ | 365/404 [00:13<00:01, 23.82it/s] Saving model.layers.28.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 91%|█████████ | 366/404 [00:14<00:01, 25.41it/s] Saving model.layers.28.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 91%|█████████ | 366/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.input_layernorm.weight (f16, [2560]): 91%|█████████ | 367/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 91%|█████████ | 368/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 91%|█████████▏| 369/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 92%|█████████▏| 370/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 92%|█████████▏| 371/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.conv1d.weight (f16, [8192,1,4]): 92%|█████████▏| 372/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.conv1d.bias (f16, [8192]): 92%|█████████▏| 373/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.A_log (f32, [32]): 93%|█████████▎| 374/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.dt_bias (f32, [32]): 93%|█████████▎| 375/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.norm.weight (f16, [128]): 93%|█████████▎| 376/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 93%|█████████▎| 377/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.post_attention_layernorm.weight (f16, [2560]): 94%|█████████▎| 378/404 [00:14<00:01, 25.41it/s] Saving model.layers.29.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 94%|█████████▍| 379/404 [00:14<00:00, 25.41it/s] Saving model.layers.29.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 94%|█████████▍| 380/404 [00:14<00:00, 26.08it/s] Saving model.layers.29.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 94%|█████████▍| 380/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.input_layernorm.weight (f16, [2560]): 94%|█████████▍| 381/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.qkv_proj.MatMul.weight (f16, [2560,8192]): 95%|█████████▍| 382/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.z_proj.MatMul.weight (f16, [2560,4096]): 95%|█████████▍| 383/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.b_proj.MatMul.weight (f16, [2560,32]): 95%|█████████▌| 384/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.a_proj.MatMul.weight (f16, [2560,32]): 95%|█████████▌| 385/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.conv1d.weight (f16, [8192,1,4]): 96%|█████████▌| 386/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.conv1d.bias (f16, [8192]): 96%|█████████▌| 387/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.A_log (f32, [32]): 96%|█████████▌| 388/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.dt_bias (f32, [32]): 96%|█████████▋| 389/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.norm.weight (f16, [128]): 97%|█████████▋| 390/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.linear_attn.out_proj.MatMul.weight (f16, [4096,2560]): 97%|█████████▋| 391/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.post_attention_layernorm.weight (f16, [2560]): 97%|█████████▋| 392/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 97%|█████████▋| 393/404 [00:14<00:00, 26.08it/s] Saving model.layers.30.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 98%|█████████▊| 394/404 [00:15<00:00, 25.98it/s] Saving model.layers.30.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 98%|█████████▊| 394/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.input_layernorm.weight (f16, [2560]): 98%|█████████▊| 395/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 98%|█████████▊| 396/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.attn.q_norm.layernorm.weight (f16, [256]): 98%|█████████▊| 397/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.attn.k_norm.layernorm.weight (f16, [256]): 99%|█████████▊| 398/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.attn.o_proj.MatMul.weight (f16, [4096,2560]): 99%|█████████▉| 399/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.post_attention_layernorm.weight (f16, [2560]): 99%|█████████▉| 400/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 99%|█████████▉| 401/404 [00:15<00:00, 25.98it/s] Saving model.layers.31.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 100%|█████████▉| 402/404 [00:15<00:00, 21.55it/s] Saving model.layers.31.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 100%|█████████▉| 402/404 [00:15<00:00, 21.55it/s] Saving model.layers.32.final_norm_layernorm.weight (f16, [2560]): 100%|█████████▉| 403/404 [00:15<00:00, 21.55it/s] Saving lm_head.MatMul.weight (f16, [2560,248320]): 100%|██████████| 404/404 [00:15<00:00, 21.55it/s] Saving lm_head.MatMul.weight (f16, [2560,248320]): 100%|██████████| 404/404 [00:17<00:00, 23.52it/s] Saving ONNX model in /home/anon/knowline/onnx/model 0it [00:00, ?it/s] Saving model.mtp.pre_fc_norm_embedding.weight (f16, [2560]): 7%|▋ | 1/15 [00:00<00:00, 678.47it/s] Saving model.mtp.pre_fc_norm_hidden.weight (f16, [2560]): 13%|█▎ | 2/15 [00:00<00:00, 1191.90it/s] Saving model.mtp.fc.MatMul.weight (f16, [5120,2560]): 20%|██ | 3/15 [00:00<00:00, 1704.08it/s] Saving model.layers.0.input_layernorm.weight (f16, [2560]): 27%|██▋ | 4/15 [00:00<00:00, 183.98it/s] Saving model.layers.0.attn.qkv_proj.MatMul.weight (f16, [2560,10240]): 33%|███▎ | 5/15 [00:00<00:00, 227.77it/s] Saving model.layers.0.attn.q_norm.layernorm.weight (f16, [256]): 40%|████ | 6/15 [00:00<00:00, 108.10it/s] Saving model.layers.0.attn.k_norm.layernorm.weight (f16, [256]): 47%|████▋ | 7/15 [00:00<00:00, 120.96it/s] Saving cos_cache (f32, [262144,32]): 53%|█████▎ | 8/15 [00:00<00:00, 137.95it/s] Saving sin_cache (f32, [262144,32]): 60%|██████ | 9/15 [00:00<00:00, 126.02it/s] Saving model.layers.0.attn.o_proj.MatMul.weight (f16, [4096,2560]): 67%|██████▋ | 10/15 [00:00<00:00, 125.93it/s] Saving model.layers.0.attn.o_proj.MatMul.weight (f16, [4096,2560]): 73%|███████▎ | 11/15 [00:00<00:00, 104.02it/s] Saving model.layers.0.post_attention_layernorm.weight (f16, [2560]): 73%|███████▎ | 11/15 [00:00<00:00, 104.02it/s] Saving model.layers.0.mlp.gate_up_proj.MatMul.input_1 (f16, [2560,18432]): 80%|████████ | 12/15 [00:00<00:00, 104.02it/s] Saving model.layers.0.mlp.down_proj.MatMul.weight (f16, [9216,2560]): 87%|████████▋ | 13/15 [00:00<00:00, 104.02it/s] Saving model.layers.1.final_norm_layernorm.weight (f16, [2560]): 93%|█████████▎| 14/15 [00:00<00:00, 104.02it/s] Saving lm_head.MatMul.weight (f16, [2560,248320]): 100%|██████████| 15/15 [00:00<00:00, 104.02it/s] Saving lm_head.MatMul.weight (f16, [2560,248320]): 100%|██████████| 15/15 [00:01<00:00, 9.50it/s] Configured 1271 MB of runtime initializer sharing; /home/anon/knowline/onnx/model/mtp.onnx.data remains self-contained. Warning: could not read generation_config.json (/home/anon/LLM/KnowLine-4B-Gen5.5 does not appear to have a file named generation_config.json. Checkout 'https://huggingface.co//home/anon/LLM/KnowLine-4B-Gen5.5/tree/main' for available files.). Falling back to config.json. Adding the tokenizer's end-of-turn token <|im_end|> (id 248046) to eos_token_id from the model configuration: 248044. Saving GenAI config in /home/anon/knowline/onnx/model Added 'mtp' section to genai_config.json Saving processing files in /home/anon/knowline/onnx/model for GenAI EXIT:0